2.0 -//Pentabarf//Schedule//EN PUBLISH DVY8SG@@cfp.balccon.org -DVY8SG Opening en en 20260918T130000 20260918T131500 0.01500 Opening Opening ceromony PUBLIC CONFIRMED Lightning talk https://cfp.balccon.org/balccon2k26-2026/talk/DVY8SG/ Tesla BalCCon PUBLISH VHWVVY@@cfp.balccon.org -VHWVVY AI can't solder or Imagine (yet) en en 20260918T131500 20260918T135500 0.04000 AI can't solder or Imagine (yet) AI has definitely changed the landscape. But while everyone is looking to find the latest and greatest prompt injection , in the forgotten back rooms there are others marching on with what they do best. Make/hack/break things. In this talk , elkentaro will talk about some of his pre-AI days creating and hacking tools and hardware the stories behind them. He will also go on and explain how AI has changed the process for him. Originally this talk was titled "I wanted to be Q from James Bond..I ended up being a hacker” but without "AI something something" its not as hot. It will be a talk about making/breaking/hacking things, mostly wireless gadgets and other obsure makings of elkentaro and how the recent rise of AI has changed some, but hasn't changed other aspects of being a hacker/maker. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/VHWVVY/ Tesla elkentaro PUBLISH RNJ3DG@@cfp.balccon.org -RNJ3DG A vision for software freedom in 2048 en en 20260918T140000 20260918T150000 1.00000 A vision for software freedom in 2048 Our litigation against Apple in front of the European Court of Justice, pushing for sustainable long term funding for Free Software in the EU and member states, "Public Money? Public Code!", Device Neutrality, Router Freedom, Free Your Android, assistance with licensing questions, a European coding competition for teenagers, and a tale of software, skateboards, and raspberry ice cream. These are some of the activities by the Free Software Foundation Europe (FSFE), which this year celebrates its 25 anniversary in empowering users to control technology. How would the world like in our area in 2048, if the FSFE has been successful? This talk will give an overview of the FSFE's vision and invite participants to give feedback on the next decades of our journey. PUBLIC CONFIRMED Talk60 https://cfp.balccon.org/balccon2k26-2026/talk/RNJ3DG/ Tesla Matthias Kirschner PUBLISH AP9VB7@@cfp.balccon.org -AP9VB7 The Agents of Chaos: AI Driven Malware Generation en en 20260918T150500 20260918T155000 0.04500 The Agents of Chaos: AI Driven Malware Generation AI has already changed the offensive security space significantly, from autonomous phishing campaigns and deepfake based social engineering to AI assisted vulnerability research and fuzzing. Malware is the next step, and the question isn’t whether AI will be used to generate it, but how far that’s already gone and where it leads. This talk covers a research project that builds an autonomous agent to generate new, functional, never seen before malware samples from scratch. The focus is on the full process: what models to use and why, how to prompt them, how to ensure the output compiles and works, how to break static signatures, and how to scale generation into the millions of unique samples. The main topics covered: AI driven offensive security and AI driven malware AI is already being used across the offensive security space, from phishing to full vulnerability research and fuzzing, and malware seems like the obvious step as AI “will replace all developers”, why not malware developers? What does AI driven malware even mean? What are the different possibilities that AI in malware gives us? The generation process: models, prompts, and workflow Before automating anything, manual testing across of models and prompts is needed. After that, how can we ensure that the malware is random, compileable, and will work flawlessly? There are several steps that need to be taken before that can be achieved: from planning the malware, to writing and fixing it, to then actually validating its functionality with AI as a judge, all without human interaction. Adding variety in the samples The agent makes a lot of decisions based on a plan that it creates beforehand, for example, there are several ways to traverse a directory in Windows, there are several encryption algorithms that can be used, all of these and more do affect the result. In restructuring the project from one file to several, adding different languages, asking for specific capabilities (without specifying how to implement them), all adding to the number of decisions and possibilities. AI malware in the wild Malicious actors are already using these ideas in the wild: from general vibe coding and assisted development to full autonomous agents and workflows that create full malware and attack frameworks, to deter and disrupt defence mechanisms and blue teams. Takeaways for defenders and attackers AI generated malware lowers the bar for attackers significantly: guardrails on frontier models are consistently bypassed, and local models require no permissions at all, but behaviour based detection remains effective because functional patterns persist even as signature changes. The same generation pipeline can be turned into a red teaming tool to test your own systems and see what you know about your environment. What Can You Gain From This • A technical walkthrough of how an autonomous agent generates functional malware samples end to end, including the prompting strategies, model selection, and loops involved. • Test results comparing different models, prompts, and methods, while sticking to actual detection rates. • A framework for generating diverse malware samples at scale for use in testing detection systems. • Documented real world examples of AI assisted and AI integrated malware from attributed threat actors. PUBLIC CONFIRMED Talk45 https://cfp.balccon.org/balccon2k26-2026/talk/AP9VB7/ Tesla Arad Donenfeld PUBLISH PQ9QZT@@cfp.balccon.org -PQ9QZT Comparing Malicious Files 2.0 en en 20260918T160000 20260918T170000 1.00000 Comparing Malicious Files 2.0 How do you use AI models on code that has to be correct, without their confident mistakes silently landing in your output? This talk answers that with a concrete, repeatable process, demonstrated end to end on a real library, and on the larger tool that library is being built for. That tool is malbench, a local-first triage and clustering workbench for malware analysts. It pulls samples and threat intelligence from various sources, computes similarity digests, runs YARA rules, and then turns a flat pile of hashes into structure: it clusters and graphs files by what they actually share and weaves several of those lenses into one view. The subject of the overall process is sdhash, a similarity-digest (fuzzy hashing) algorithm that fingerprints binary data into bloom filters and scores two fingerprints for similarity. It has a C++ reference by Vassil Roussev and Candice Quates and a 2021 Go port that had gone dormant for over three years. This is exactly the kind of valuable-but-abandoned code worth adopting rather than rewriting. The goal was to bring it forward into a clean, modern, dependency-free Go library with an idiomatic sealed API, full documentation, and total statement coverage. The spine of the effort, and the core defense against hallucination, is that ground truth comes first. Before trusting a single line the model produced, I built a deterministic, seed-reproducible corpus and parallel C++/Go harnesses, ran it through both, and diffed the outputs with an independent tool. That turns the reference into an oracle: any divergence, a real bug or something a model invented, surfaces immediately as a mismatch. Generation parity held across more than a hundred thousand files and scoring parity across nearly three million pair comparisons in both modes, at zero unexplained divergences. The process surfaced first three correctness problems in the C++ implementation scoring as well as a problem with the hashing algorithm itself. Each change to the algorithm was carefully isolated and the effects on scoring measured. This talk goes into detail all of the tooling and process used to get sdhash up to snuff and then plug it in to malbench. You will also learn about the graphing and clustering algorithms used to visualize the results of the different malware hashes. PUBLIC CONFIRMED Talk60 https://cfp.balccon.org/balccon2k26-2026/talk/PQ9QZT/ Tesla Malware Utkonos PUBLISH BBTUHZ@@cfp.balccon.org -BBTUHZ Hacking in the Middle of the Ocean en en 20260918T170000 20260918T173000 0.03000 Hacking in the Middle of the Ocean Talk will show some unique challenges as well as speaker vulnerability research in maritime industry. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/BBTUHZ/ Tesla Vlatko Kosturjak PUBLISH UEYM89@@cfp.balccon.org -UEYM89 Hunting for business logic vulnerabilities en en 20260918T173000 20260918T183000 1.00000 Hunting for business logic vulnerabilities There are a number of well known classes of vulnerabilities that an enterprising hacker or penetration tester wants to uncover in an application. Some of those issues are harder to detect or exploit because of well implemented browser security mechanisms or because of various improvements in web application frameworks that hide the potentially dangerous methods from the developers. In addition to the points outlined above, the collective awareness about common security issues, vulnerabilities and potential weaknesses has been raised, making discovery of potential issues more difficult, especially in hard, well audited targets. In such cases, instead of subverting the code flow, an attacker might try to subvert the applications logic or even better manipulate the business process that the application supports. This class of vulnerabilities is commonly referred to as business logic vulnerabilities, and when discovered in the wild and reported, all the specific and different nuanced cases of vulnerabilities are usually thrown into the bucket labeled "business logic" vulnerabilities. But when we review such issues, we can see that each case is unique. This talk will present the result of a research study where the author manually reviewed about 300 publicly disclosed vulnerability reports and tried to classify and cluster discovered vulnerabilities into a few categories that can be used to detect business logic issues in applications. So let's take a ride through some real life cases and examples on how to manipulate calculation, assumptions, processes, branching, logical and time based TOCTOU and other fun cases on how to break an application. PUBLIC CONFIRMED Talk60 https://cfp.balccon.org/balccon2k26-2026/talk/UEYM89/ Tesla Tonimir Kisasondi PUBLISH QF7RUJ@@cfp.balccon.org -QF7RUJ You build vulnerable hardware accidentally. I do it on purpose. We are not the same. (Behind the scenes of building hardware CTF challenges) en en 20260918T183000 20260918T190000 0.03000 You build vulnerable hardware accidentally. I do it on purpose. We are not the same. (Behind the scenes of building hardware CTF challenges) Our talk presents the process of creating the hardware challenges used in the Serbian National CTF in 2025 and 2026. In both editions, we designed custom embedded devices with intentionally introduced vulnerabilities, built specifically to be exploited during the competition. From selecting realistic attack vectors and hardware components, designing PCBs and firmware, manufacturing the boards, and preparing the competition environment, to watching students solve them during the event, we were involved in every stage of the process. This talk explores how the challenges are conceived, designed, built, and deployed. Along the way, we will discuss the technical and practical trade-offs, the lessons we learned, and the challenges of creating educational, engaging, and reliable hardware CTF tasks. The talk is suitable for attendees with a beginner/intermediate level of technical knowledge. Whether one is interested in hardware security, embedded systems, or CTF competitions, they will gain a behind-the-scenes understanding of how hardware challenges are designed, built, and deployed. No prior experience with hardware hacking or specialized tools is required. The talk will follow the outline below: ### Motivation - Why we decided to introduce hardware challenges to the Serbian national CTF. - Inspiration from ECSC, where hardware challenges have been a regular competition category. - Our backgrounds and how the project came together. ### Hardware CTF Design Principles - What makes a good hardware CTF challenge. - Constraints and trade-offs: budget, accessibility, educational value, realism, and fun. - Examples of hardware challenges from ECSC and the ideas that inspired our designs. ### 2025 Challenge - Initial concept and design goals. - Hardware architecture, component selection, and PCB design. - Manufacturing. - Challenge narrative and participant experience. - Vulnerabilities and intended attack paths: - UART - eFuse - USB HID - Vulnerable OTA updates - Unsafe cryptographic secret storage - Reverse engineering ### 2026 Challenge - Design goals and concept of a fictional game console. - Hardware architecture, component selection, and PCB design. - Manufacturing. - Challenge narrative and gameplay. - Vulnerabilities and attack techniques: - Introductory side-channel analysis - SPI bus sniffing and display reconstruction - Logic analyzer usage - Timing attacks using PIO ### Lessons Learned - Challenge balancing and playtesting. - Designing intentional vulnerabilities. - Manufacturing, logistics, and deployment during the competition. - What worked well and what we would do differently. ### Future Directions Ideas for future competition challenges: - Glitching and fault injection. - Power analysis. - NFC/RFID. - Radio protocols. - CAN bus. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/QF7RUJ/ Tesla Maja Miljanić TheProxy PUBLISH AKXHPM@@cfp.balccon.org -AKXHPM Your Lock(er) Knows Your PIN ... And So Do I en en 20260918T190000 20260918T194500 0.04500 Your Lock(er) Knows Your PIN ... And So Do I This talk is a continuation of our DEFCON 32 research on electronic locker locks. We discuss the general problem of how electronic locks handle user-chosen secrets, revisiting vulnerabilities in locks from multiple manufacturers. We focus on what these devices store: user PINs, RFID UIDs, manager credentials, and audit logs - often in plaintext and trivially extractable. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/AKXHPM/ Tesla Dennis Giese PUBLISH 8HJGZ8@@cfp.balccon.org -8HJGZ8 Fun with virtualization en en 20260918T194500 20260918T204500 1.00000 Fun with virtualization Working in Incident Response sometimes means you get to go on a grand adventure, where you run into adversaries employing clever techniques to make your digital firefighting life difficult. In this talk we take you along on our journey, cover the People, Processes and Technologies angles and hopefully give you some tips and tricks for dealing with these types of threats. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/8HJGZ8/ Tesla Hank Scorpio Scorpio Hank PUBLISH ZKQTEK@@cfp.balccon.org -ZKQTEK Hacker Jeopardy en en 20260918T223000 20260919T013000 3.00000 Hacker Jeopardy We will play some rounds of hacker jeopardy. If you ask yourself: "What the heck is Jeopardy?", you already got the gist of the game, as all answers need to be formulated as questions! Participants will have to answer questions in different categories to get the most points. But beware, it's not only about knowing the answer, you also need to be fast with a buzzer! Are you up for the challenge? PUBLIC CONFIRMED Workshop120 https://cfp.balccon.org/balccon2k26-2026/talk/ZKQTEK/ Tesla cluosh georg Hetti PUBLISH 78CRUA@@cfp.balccon.org -78CRUA FROST: SSD Side Channels from the Browser, and Why You Should Care en en 20260918T140000 20260918T143000 0.03000 FROST: SSD Side Channels from the Browser, and Why You Should Care Earlier this year, we published "FROST: Fingerprinting Remotely using OPFS-Based SSD Timing". The paper made the news, likely because "a website can spy on you if you click a link" makes a good headline. Reactions were split: Some people wanted to disable JavaScript completely, claiming that allowing websites to execute scripts on client devices was a mistake in the first place. Others waved the attack off as a toy example that only works in a lab, never feasible in the real world. For us, the truth lies somewhere in the middle. FROST is a real attack. From plain JavaScript, we measure SSD contention from the browser and use it to fingerprint the websites you visit and the apps you open, without requiring any additional interaction beyond clicking a malicious link. It's also fragile: Classification depends on training data, and prior work has shown that trained models cannot easily be generalized to different SSD models. This talk presents the attack, and gives some intuition about the underlying insights. Does the average user need to be scared? Who actually needs to care? Why do browsers allow this in the first place? And what are side-channel attacks anyways? Additionally, we give an overview of other remote side-channel attacks, showing what's possible in theory to a motivated attacker. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/78CRUA/ Pupin Hannes Weissteiner PUBLISH TVDMBV@@cfp.balccon.org -TVDMBV Detecting Linux rootkits: Know where to look in user-space en en 20260918T143000 20260918T150000 0.03000 Detecting Linux rootkits: Know where to look in user-space Rootkits are designed to hide themselves or other malicious software from users' and administrators' prying eyes. To create the illusion that nothing is there and everybody is fine to move along, they usually subvert tool output, standard library functions or kernel system calls to hide the presence of specific processes or files. This illusion is often just convincing enough to fool standard tools, but if we put a little more effort into observing system behavior, we can still see shadows of what has been hidden, by relying on traces that are hard to cover by rootkit authors. In this talk I take a look at rootkit implementations that subvert the system at different layers – the system call interface, the standard library, or through eBPF probes. I give an overview over detection techniques that have been implemented in traditional rootkit hunting scripts, plus a few novel methods. I present a modern implementation of the most promising techniques that can be integrated with existing live-forensic capabilities to hunt for rootkits at scale. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/TVDMBV/ Pupin Hilko Bengen PUBLISH ZXZZQN@@cfp.balccon.org -ZXZZQN Deconstructing Modern macOS Initial Access Vectors en en 20260918T150000 20260918T154000 0.04000 Deconstructing Modern macOS Initial Access Vectors We begin by exploring the top of the funnel. Attackers have moved far beyond easily identifiable spam. We will deconstruct recent campaigns to show how adversaries are successfully lowering user guard through: How threat actors weaponize Google Ads to push malicious software disguised as legitimate enterprise tools (e.g., Slack, Notion, or VPN clients). A deep dive into localized, highly convincing fake browser updates and system notification campaigns that socially engineer users into bypassing native warnings. Once the user interacts with the lure, how does the malware actually run? macOS is a rich Unix-based environment with multiple scripting avenues. We will analyze the "Living off the Land" (LotL) techniques currently dominating the macOS threat landscape, including: The use of Bash, Zsh, and legacy Python/Perl scripts to establish persistence and pull down secondary payloads. How attackers weaponize Apple’s native automation languages to silently interact with system APIs, bypass sandbox restrictions, and generate convincing fake credential prompts. Techniques used by threat actors to obfuscate their code, making static analysis incredibly difficult for defenders. Finally, we will break down how these threats are packaged to evade Gatekeeper and initial static analysis. We will compare and contrast real-world samples across: The weaponization of standard Apple Disk Images (.dmg) and Installer Packages (.pkg), including pre-install/post-install script abuse. The shift from standalone Mach-O binaries to hiding malicious routines inside Platypus-packaged applications and bloated Electron frameworks, which are notoriously difficult for traditional AV to parse effectively. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/ZXZZQN/ Pupin Stephan Berger PUBLISH WNJR8M@@cfp.balccon.org -WNJR8M Renting Brains, Owning the Mistakes: LLMs in Cybersecurity Education en en 20260918T160000 20260918T170000 1.00000 Renting Brains, Owning the Mistakes: LLMs in Cybersecurity Education Total length : 45-50 min + 10-15 min Q&A Part 0 · Who, and why listen whoami: PhD researcher, occasional lecturer, TurkuSec chair; teaching cybersecurity at the University of Turku. Brief mention UTU / TurkuSec context for credibility. Core idea is that an LLM is neither enemy nor friend a tool with a user, and the user owns the mistakes. ------------------------------------------------------------------------------------------------------------------------------------------------------------------- Part 1 · Why we care Evangelist people think LLMs are good, and get only better, and it is impossible to get harm from it, and while they measure who burnt more tokens within 24h frames, people pay for it, and sometimes they pay too much. "It's cheap" → economics. Uber burned its planned 2026 AI-coding budget in four months; engineers at $500–$2,000/mo; OpenClaw chewing $1–5k/day on a $200 plan; GitHub freezing Copilot sign-ups. At today's subsidized pricing the unit economics are propped up, not "it will collapse," but someone else is absorbing the bill. "It's eco-friendly" → energy & where it lands. Tiny per prompt (0.24 Wh) vs vast in aggregate (~945 TWh by 2030); Jevons paradox; Google's own emissions up despite efficiency gains → then the local cost: Vantage VA, xAI Memphis, and externalities on bills, rent, land, sleep. LLM data centers cause severe pollution and harm people. "It's a revolution, not a bubble" → jobs & failures. Layoffs framed as AI efficiency (Oracle, Meta, Microsoft, Amazon); real-world breakage (AWS/Kiro outage, the wiped DataTalks database, the Fastly senior-vs-junior split). These claims are backed by independent papers and expert review of the waste and pollution LLMs produce. Other claims supporting the point are interviews of local people complaining about noise, pollution and general detrimental impact of LLMs. Students use these tools constantly and won't stop at graduation. The habits formed in the lecture hall walk into the SOC. So this is an education problem — which is where the rest of the talk lives. ------------------------------------------------------------------------------------------------------------------------------------------------------------------- Part 2 · The new learning reality Taking Part 1 into account, why don't we want students to abuse LLMs? What they do at University becomes a habit once they start working. Students already use LLMs for everything (to explain, to summarize, to write, to code, to debug, to exam-prep, and sometimes to skip the learning entirely). UTU permits responsible use; the question is no longer whether but how. Google-fu is dying from "find and think" to "ask and accept" (StackOverflow decline; same effect we observe at university, students do not google basic problems, and when LLM troubleshooting fails they immediately email us, and we reply with the first link on google). The real problem: usage without structure, without knowing what these tools are good at, bad at, and where they quietly fail. ------------------------------------------------------------------------------------------------------------------------------------------------------------------- Part 3 · Where it quietly erodes — the bad Faster learning, weaker thinking: outsourcing the struggle ≠ growth; a correct output doesn't prove understanding. Most students don't verify; fluency is mistaken for correctness; beginners are most exposed to smooth nonsense. Since most of the students LACK the experience (especially relevant for bachelor students), they immediately believe LLM, even when the facts are wrong. Hands-on skills vs AI assistance: learn the underlying skill before automating it. Some students lack basic IT skills, and LLM is making it worse. The purpose of homework is to solidify the knowledge on the matter, not to feed it to LLM for training the model and getting your answers. Sharing a couple of negative examples from teaching time at UTU (no text on slide, but rather storytelling, 3 mins) Sharing students' feedback on LLM incorporation to the course (This is being collected currently, ready in August) ------------------------------------------------------------------------------------------------------------------------------------------------------------------- Part 4 · The flip: where it genuinely helps — the good The pivot: same tool, different user. The fluent output that fools a beginner teaches a careful student what good looks like. Personal tutor: patient, available at 2am, removes the social cost of asking — especially for shy, Finnish, and non-native students. From theory to "it works": environment-specific debugging companionship; the moment students used to quit becomes the moment they get unstuck. It generates a lot of research avenues → master's/bachelor's theses, PhD dissertations, research assistants. Force multiplier for instructors: faster lab design, exam variants, tighter feedback loops. Self-study that finally works; serious entry points beyond the syllabus. Language equity: non-native speakers judged on their security thinking, not their prepositions. Sharing a couple of positive examples from teaching time at UTU (no text on slide, but rather storytelling, 5-6 mins) Sharing the students' feedback on LLM incorporation to the course (This is being collected currently, ready in August) ------------------------------------------------------------------------------------------------------------------------------------------------------------------- Part 5 · Why verification is the whole game — capstone risk The hardest risk we are afraid of at UTU isn't hallucination, it's bias you can't see. Historical analogues: Sugar Research Foundation / Harvard (1967); Coca-Cola's GEBN; Merck/Vioxx ghostwriting (~55k deaths, Graham's FDA testimony). Those manipulations had to clear high bars and fool trained audiences. The audience for LLMs is everyone — including future doctors and regulators, while they're still students. The 2+2=5 problem: we learned to question the press and social media; we haven't learned to question the model. The presentation is the persuasion. What if tomorrow an LLM starts telling students that telnet is a good idea or introduces some subtler bias? Because students trust the model, they won't double-check the claim; they'll simply believe it. ------------------------------------------------------------------------------------------------------------------------------------------------------------------- Part 6 · Takeaways & close Neither enemy nor friend; a tool with a user, but the current approach is alarming Verification is the new baseline literacy. Compress work, don't replace judgment. Cybersecurity education is uniquely positioned to lead. The students who benefit most are the ones we were quietly losing. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/WNJR8M/ Pupin Ismayil PUBLISH KYZCRH@@cfp.balccon.org -KYZCRH Human Error Is Not the Problem: How Hiring, Culture and Psychology Shape Cyber Risk en en 20260918T170000 20260918T173000 0.03000 Human Error Is Not the Problem: How Hiring, Culture and Psychology Shape Cyber Risk 1. Introduction: why “human error” is an incomplete explanation The talk starts by challenging the usual narrative that people are simply careless, lazy or untrained. In many cases, insecure behavior is a predictable result of the environment: pressure, unclear responsibilities, weak processes, bad communication and poor hiring decisions. 2. Where cyber risk begins before the cyber team sees it This part explains how risk enters the organization through recruitment, onboarding, role design, access decisions and organizational culture. Examples include rushed hiring, unverified candidates, fake profiles, poor reference checking, excessive access, lack of psychological safety and unclear escalation paths. 3. Social engineering and the psychology of trust The talk explores why people trust the wrong signals: authority, urgency, familiarity, similarity, politeness and fear of conflict. It connects phishing, impersonation, fake candidates and internal manipulation to basic psychological mechanisms. 4. Insider risk without Hollywood drama This section explains that insider risk is not only malicious employees stealing data. It can also include frustrated employees, overloaded teams, people bypassing procedures, unmanaged contractors, unclear accountability and people with access they no longer need. 5. Why awareness training is not enough Security awareness often fails because it treats people as isolated decision-makers. The talk explains why behavior changes only when incentives, workflows, leadership behavior and consequences are aligned. 6. What HR and security teams should do together The final practical section suggests a basic cooperation model between HR, security and leadership: better hiring checks, access hygiene, role-based onboarding, psychological safety for reporting, exit procedures, manager training and clearer internal communication. 7. Conclusion The talk ends with a simple message: people are not the weakest link by default. Poorly designed systems make them weak. Better human systems create better security. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/KYZCRH/ Pupin Nataša Vasić PUBLISH MPPFNF@@cfp.balccon.org -MPPFNF The Hitchhikers Guide to Hacking Cheap Bluetooth Speakers en en 20260918T173000 20260918T183000 1.00000 The Hitchhikers Guide to Hacking Cheap Bluetooth Speakers Deep inside the shelves of our nearest IKEA store, hidden between smart appliances, there lies the cheap IKEA KALLSUP, a Bluetooth speaker, that promises limitless possibilities. Well, maybe not limitless. But it does allow synchronization of up to 100 devices. Supposedly at least, we did not buy 100 of them. However, it is not this promise of connectivity that caught our attention, but the curiosity about the intricacies of the internals of such a cheap device. Follow us along on our journey into the unknown, hitchhiking on the experiences made by those who came before us, while we dive into reverse engineering of cheap, undocumented chips, analyzing firmware, and other general tomfoolery. PUBLIC CONFIRMED Talk60 https://cfp.balccon.org/balccon2k26-2026/talk/MPPFNF/ Pupin georg cluosh PUBLISH BZXNHH@@cfp.balccon.org -BZXNHH A Street Sign, a Shadow, and an Answer: OSINT Workshop en en 20260918T183000 20260918T203000 2.00000 A Street Sign, a Shadow, and an Answer: OSINT Workshop The workshop is split across two conference days. Day one is a practical introduction to investigative methodology through live demonstrations and guided walkthroughs. Rather than memorising tools, participants learn how to think like investigators: building hypotheses, verifying evidence, eliminating bad assumptions, and documenting conclusions that others can reproduce. Topics include geolocation, image verification, archive research, metadata, social media investigation, and practical search techniques, all demonstrated using publicly available information. Day two is a competitive OSINT CTF/Hackathon where participants apply those techniques to solve a series of progressively harder challenges inspired by real investigations. Teams of up to three compete to identify people, places, events, and relationships using only open sources. Challenges reward both speed and investigative rigour, with recognition for the fastest verified solutions as well as the most elegant investigative process. After spending several hours forcing ambiguous information into defensible answers, participants often find themselves applying the same methodology everywhere else, from incident response and threat intelligence to random travel photos and breach data. The goal is not simply to teach OSINT, but to develop a way of approaching problems that remains useful long after the workshop ends. PUBLIC CONFIRMED Workshop https://cfp.balccon.org/balccon2k26-2026/talk/BZXNHH/ Pupin Jurica Radović PUBLISH USMV9H@@cfp.balccon.org -USMV9H From Zero to root in 120 minutes - Introduction to Wordpress Hacking en en 20260918T203000 20260918T223000 2.00000 From Zero to root in 120 minutes - Introduction to Wordpress Hacking This is an introductionary level workshop targeted at a novice/beginner level audience that wants to learn how "hacking" actually works. InfoSec personel and other "professionals" attending this session will get shanghaied into supporting the other attendees. Prerequisites: Bring your own/a Laptop running a recent version of Kali-Linux inside a virtual machine or from a USB stick. PUBLIC CONFIRMED Workshop120 https://cfp.balccon.org/balccon2k26-2026/talk/USMV9H/ Pupin leyrer PUBLISH JRDFXS@@cfp.balccon.org -JRDFXS Karaoke - Some sing to remember, some sing to forget en en 20260918T220000 20260919T090600 11.00600 Karaoke - Some sing to remember, some sing to forget People attend. People sing. Everyone has a good time. PUBLIC CONFIRMED Workshop120 https://cfp.balccon.org/balccon2k26-2026/talk/JRDFXS/ Lounge MacLemon PUBLISH UZQQAG@@cfp.balccon.org -UZQQAG Digital Oncologists Require Cyber Care - Securing AI Agents in Radiotherapy en en 20260919T110000 20260919T115000 0.05000 Digital Oncologists Require Cyber Care - Securing AI Agents in Radiotherapy - Cancer care enabled by AI - MITRE ATLAS & SAFE-AI Crash Course - Case Study: Radiotherapy Management Agent - From Threat Model to Pentest Report PUBLIC CONFIRMED Talk60 https://cfp.balccon.org/balccon2k26-2026/talk/UZQQAG/ Tesla Jani Kovacs PUBLISH EP88CX@@cfp.balccon.org -EP88CX REGOČ my SW/HW AI team en en 20260919T120000 20260919T124500 0.04500 REGOČ my SW/HW AI team Over the past year I've been using REGOČ — my team of 11 AI agents — to work on real hardware projects, not just chat demos. We ported the legendary **PDP-1 to the ULX3S FPGA**, try to brought up the open-source **GateMate FPGA Ethernet** stack with the KSZ9031 PHY, and reverse-engineered the **Chasing Gladius Pro underwater drone** — sniffing MAVLink, RTP video and the WiFi handshake to build our own control PWA on a Raspberry Pi. On the simulation side, we built **FTSIM**, an openEMS-based pipeline that runs FDTD signal-integrity checks on real PCB Gerbers across seven different boards. We also built a small **EMC pre-certification UI** for our in-house lab, and used the agents as a permanent "second pair of eyes" for **KiCad** — catching footprint mistakes, validating diff-pair routing, generating Gerber exports and fixing weird stackup issues before they hit production. PUBLIC CONFIRMED Talk45 https://cfp.balccon.org/balccon2k26-2026/talk/EP88CX/ Tesla Goran PUBLISH JYK3BH@@cfp.balccon.org -JYK3BH Universal Plug and Pwn en en 20260919T124500 20260919T133000 0.04500 Universal Plug and Pwn The talk will cover the following topics: - Opening up the devices - Dumping the Firmware - Firmware Reverse Engineering - Identifying interesting entry points - Vulnerabilities we uncovered - Mitigations - Keeping your own devices safe This talk is suitable for beginners. We will show how to start analyzing your own devices, and how easy it is to exploit devices where security was not a priority during development. PUBLIC CONFIRMED Talk45 https://cfp.balccon.org/balccon2k26-2026/talk/JYK3BH/ Tesla Andreas Markus Kevin PUBLISH J8S7WP@@cfp.balccon.org -J8S7WP Cryptography with Serbian eID Cards en en 20260919T133000 20260919T143000 1.00000 Cryptography with Serbian eID Cards 60-minute lecture in which you will learn more than you ever wanted to know about Serbian eID cards PUBLIC CONFIRMED Talk60 https://cfp.balccon.org/balccon2k26-2026/talk/J8S7WP/ Tesla Nikola Ubavić PUBLISH REHJAL@@cfp.balccon.org -REHJAL Why you shouldn’t worry about your SAP systems… or should you? en en 20260919T143000 20260919T150000 0.03000 Why you shouldn’t worry about your SAP systems… or should you? The talk will include the following aspects: 1. Understanding SAP in the Enterprise Landscape: - what SAP is and its role in large organisations - overview of the most widely used SAP products by the business, IT and OT - business processes typically managed by SAP systems. 2. SAP Architecture and Deployment Models: - common SAP deployment models and security considerations - common integrations with core enterprise systems and potential pivoting possibilities - typical trust relationships and common attack surfaces. 3. Security Challenges and Vulnerability Trends: - the most common SAP vulnerabilities and misconfigurations - recent security trends and attack techniques - why SAP environments are often a blind spot for SOC and security teams. 4. Improving SAP Security and Visibility - protection strategies for SAP environments - detection and monitoring approaches for SOC teams - immediate actions to strengthen SAP security posture. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/REHJAL/ Tesla Anita Cwynar PUBLISH PX8LSH@@cfp.balccon.org -PX8LSH Compression, how does it even work? en en 20260919T150000 20260919T160000 1.00000 Compression, how does it even work? This talk is part of my foundational technologies arc, which includes these topics: (In alphabetical order.) - Backups - Colour - Email Systems - Emoji - Encoding - Fonts and Typography - HTTP/2 - USB (2 talks) - various shell tools, including `ssh` and `tmux` PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/PX8LSH/ Tesla MacLemon PUBLISH E7GFPG@@cfp.balccon.org -E7GFPG Reverse Engineering FERMAX: Detour, Dead End, and Scope Creep en en 20260919T160000 20260919T170000 1.00000 Reverse Engineering FERMAX: Detour, Dead End, and Scope Creep When reverse engineering the proprietary DUOX PLUS intercom system dubbed the ‘most secure in world’ by FERMAX, previously Kirils & friends focused on its digital 2-wire signalling and employed such tools like oscilloscopes, logic analyzers and breadboards. While these attacks are important as they shine light on the internal workings on the system, their application in the field is limited as one would need to acquire access to the 2-wire bus, which is only possible from the inside of the building. Then we noticed something that was right in front of our eyes, Access control panels! These things are out there just on the perimeter. And, when installed on multi-tenant buildings, they have RFID reader modules installed. FERMAX offers modules doing EM4100, MIFARE Classic, and MIFARE Desfire. Even more they offer standalone Bluetooth modules too! In this talk we give an overview of previous research and expand on it by exploring the possibilities of entering the perimeter by attacking the bluetooth and RFID dimension of these systems, and exploring card cloning, implanting, and cryptographic attacks together with Iceman. In our research we extracted firmware and analyzed two different mobile applications to control the system, TUYA and NearKey. Attendees will gain insight into decoding and interacting with closed digital protocols, exposing vulnerabilities in real-world access control systems. They also get practically applying RFID attacks to real world systems in use right now. PUBLIC CONFIRMED Talk60 https://cfp.balccon.org/balccon2k26-2026/talk/E7GFPG/ Tesla Kirils Solovjovs Iceman PUBLISH UHEUPT@@cfp.balccon.org -UHEUPT Keeping Trains on Track - A Glimpse into Germany’s Railway Infrastructure en en 20260919T170000 20260919T174000 0.04000 Keeping Trains on Track - A Glimpse into Germany’s Railway Infrastructure My rough idea for the talk would be as follows: 1. What Train Infrastructure^C Dreams are Made Of: The Basic Building Blocks of Railway Infrastructure - Track Switches - Signals - Balises - Track Vacancy Detection - Hot Box Detection Systems 2. Why it is hard to Crash a Train: Interlocking Systems and Safety Logic - Different generations of interlocking systems - Safety principles 3. Train Control Systems: Why Trains do not need Speeding Tickets - Intermittent train control - Continuous train control - ETCS (European Train Control System) 4. Long live GSM(-R): The Mobile Network behind Railway Operations - Details on the relevance and functionality of GSM-R (GSM for Rail) - Differences between GSM and GSM-R 5. Spawning Wireshark: Examples of Bits and Bytes in the Wild - RaSTA - SAHARA - SBS - SCI-Protocols 6. Selected Examples of Developments to come - Automated train operation - FRMCS 7. Closing Thoughts PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/UHEUPT/ Tesla BlackCat PUBLISH KX7ULU@@cfp.balccon.org -KX7ULU Every ride you take - Hacking a City’s Public Transportation en en 20260919T174000 20260919T182000 0.04000 Every ride you take - Hacking a City’s Public Transportation Description ======= The talk is divided into 12 stages, showing step-by-step attack chains. In the first stages, I’ll relate the origin of the idea, provide information about the company, recon, and exposed .git directories that revealed source code, hardcoded credentials, and references to internal services and repositories, as well as some SQLi. In Stage 3, I’ll present a Windows-based DVR system deployed on urban buses. Using a user with almost no permissions, I obtained an LFI. Using information from vendor manuals, documentation screenshots, and recovered internal paths, .frm and .ibd files were obtained and reconstructed in a Docker container. This allowed the recovery of credentials, leading to full administrative access to the DVR platform (DEMO). With this access, it was possible to manage users and drivers, access sensitive internal data, and remotely view and control cameras and microphones installed inside buses. After some time, I figured out that this system was also used in 11 provinces, turning a local issue into a nationwide one. In Stage 6, the research returns to the exposed .git files, where a private GitLab URL was found. From there, access to CI/CD pipelines, cron jobs, and a Docker registry was obtained. While analyzing the cron jobs, multiple hardcoded credentials were found, including access to banking-related services and FTP servers. By analyzing the code and the .lock files, it was possible to upload a file to the FTP server, which was later pulled and executed by internal processes, resulting in RCE on the server. Stage 8 marks the full infrastructure compromise. Using the reverse shell, it was possible to enumerate internal services, access production databases, bypass network segmentation, and identify additional systems. Within the bus administration platform, access was obtained to the backbone of public transportation, managing drivers, users, companies, ticketing systems, real-time vehicle tracking, and remote operational controls such as fuel cutoff. Unlike the DVR platform, this system operates at city, regional, and national scale. In Stage 10, I’ll show other systems that got access, for example: taxi applications and a large-scale bike rental system, affecting approximately 710 taxi drivers and more than 115,000 registered bike service users. The final system explored was a government-related server containing highly sensitive information, including driver licenses, identification numbers, addresses, phone numbers, and operational records for taxi, bus, and private transport drivers. This was the point where I decided to stop the investigation and report it immediately. The talk ends with the responsible disclosure process, challenges encountered when reporting vulnerabilities across multiple organizations and public entities, and lessons learned about securing critical infrastructure. The goal is to show how chaining basic vulnerabilities can lead to systemic compromise, and why public mobility systems deserve the same security attention as traditional critical infrastructure. Also I would like to encourage new generations to do ethical hacking and help build stronger relationships between hackers and companies. Outline ======= - Stage 0 - Whoami - Disclaimer - Introduction - Stage 1 - Landing page - .git folders enumeration - SQLi - Stage 2: Exposed .git - Hardcoded creds - Internal services - GitLab url in .git/config - Stage 3: DVR System - Intro - Internal paths discovery - LFI - Stage 4: DVR priv escalation - LFI to get .frm/.ibd - DB recovery (DEMO) - Stage 5: DVR Admin access - User and drivers data exposure - Internal system data - Access to all cameras & microphones - National presence - Stage 6: GitLab repos & CI/CD - Public repos - Public pipelines - Docker registry exposure - Stage 7: Cron + FTP - Cron job download file from FTP - Shell upload to the FTP server - .lock execution control - Stage 8: Reverse shell - Internal network access - Full database access - New systems discovered - Stage 9: Bus administration system - Access to users/drivers/company data - Ticketing & balance management - City+Regional+National buses - Control the buses remotely - Stage 10: Additional systems - APKs - Taxi apps - Bike rental - Full administrative access - Stage 11: Gov-related server - Taxi, buses and private transport records - Driver license, IDs, address, phones, etc - Stage 12: Ending - Disclosure process - Challenges & responses - Conclusions - Q&A PUBLIC CONFIRMED Talk45 https://cfp.balccon.org/balccon2k26-2026/talk/KX7ULU/ Tesla Ignacio Navarro PUBLISH W3ARRY@@cfp.balccon.org -W3ARRY Power Analysis Attacks 101: From Waveform to Private Key en en 20260919T182000 20260919T190500 0.04500 Power Analysis Attacks 101: From Waveform to Private Key This talk covers the basics of static and differential/correlation power analysis: the methods, the math (with intuitive explanations), and, if the demo gods are feeling generous, a live key extraction on stage. No prior knowledge assumed. PUBLIC CONFIRMED Talk60 https://cfp.balccon.org/balccon2k26-2026/talk/W3ARRY/ Tesla igor PUBLISH BKDUMV@@cfp.balccon.org -BKDUMV Democratizing the creation of video tools with Open Source: Grassroots Community building and the recurBoy en en 20260919T190500 20260919T192000 0.01500 Democratizing the creation of video tools with Open Source: Grassroots Community building and the recurBoy This talk is not a technical talk, it is following an open source project from its beginnings to its evolution 7 years later and how it has grown organically in the underground video art community without the interference of its creators. Technical information about the project: recurBOY is a stand-alone digital video synthesizer and sampler. it can trigger clips and run shaders to create and manipulate sd video. outputs sd video over composite or hdmi 2 source modes to generate video : sampler and shaders process any source with additional FX control shader/fx parameters directly with 4x knobs or externally with 4x cv inputs all inputs also controllable via usb-midi process external video through compatible usb capture cards / web-cams More complete information can be found at: https://github.com/cyberboy666/recurBOY PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/BKDUMV/ Tesla Guergana PUBLISH SVGXZE@@cfp.balccon.org -SVGXZE Manufacturing Minesweeper! en en 20260919T193000 20260919T203000 1.00000 Manufacturing Minesweeper! You're a hardware hacker that likes to build devices the way they SHOULD be built - meaning, the way you like them! You've made some cool prototypes that are useful for you, so maybe they might be useful to other people too? Maybe you should manufacture some? You might help people out and even make a bit of money on the side. But the product manufacturing landscape is a figurative minefield, littered with traps that will explode your costs and time. Using real examples, I'll present all my own naive, ignorant and just plain stupid mistakes so maybe you can avoid them. PUBLIC CONFIRMED Talk60 https://cfp.balccon.org/balccon2k26-2026/talk/SVGXZE/ Tesla Zoz PUBLISH ALV7BZ@@cfp.balccon.org -ALV7BZ Lightning talks en en 20260919T203000 20260919T223000 2.00000 Lightning talks On BalCCon2k26, we want everyone to have an opportunity to speak! So we are soliciting short, but engaging 5 minute talks – Lightning Talks – from any and all attendees. The Lightning Talk format provides very personal, concise thoughts, ideas or calls for action. Typically speakers use lightning talks to talk about their experience, their opinions or for a wake-up call towards the agile community and markets. Format Everyone can speak at Balccon! Tell us about your project, idea, plans or your best jokes, just make sure you have a slide deck and keep it under 5 minutes! Proposal Now it’s your turn! Send an email at orga (at) balccon.org until 19th September 4pm with Subject: Lightning Talk - BalCCon2k26 Name: Language of your Presentation Keywords: Abstract: Relevant Links (Project Page, etc.) PUBLIC CONFIRMED Workshop120 https://cfp.balccon.org/balccon2k26-2026/talk/ALV7BZ/ Tesla BalCCon PUBLISH NJFLGX@@cfp.balccon.org -NJFLGX Defending LLMs with LLMs: A Multi-Agent Approach to Prompt Injection en en 20260919T110000 20260919T112000 0.02000 Defending LLMs with LLMs: A Multi-Agent Approach to Prompt Injection A problem that will be bigger and bigger in the future is that public-facing LLM applications (customer support bots, in-product assistants, RAG-backed chat) are now production infrastructure for thousands of companies, yet the security tooling around them is roughly where web app firewalls were in 2005: regex blocklists and single-shot classifiers. Real-world attacks (DAN-style jailbreaks, indirect injection via poisoned documents, encoding bypasses (base64, Unicode homoglyphs, leetspeak), multi-turn priming attacks) defeat these defenses routinely. Commercial guardrails (Lakera, NeMo, LLM-Guard) help but operate on a single-pass pipeline that struggles with novel, composed, or low-signal attacks. Proposed approach. I've built and open-sourced an agentic guardrail system that treats LLM defense as a multi-agent reasoning problem rather than a classification problem. The architecture runs concurrent specialist agents in a LangGraph fan-out/fan-in pattern: Injection Agent — looks at instruction-override patterns, role confusion, system prompt extraction attempts Encoding Agent — detects base64, hex, Unicode obfuscation, and language-switch bypasses Context Agent — analyzes indirect injection via retrieved documents and tool outputs Output Protection Agent — scans LLM responses for system prompt leakage, sensitive data, and policy violations PII Agent — Presidio-backed entity detection on both prompts and responses A fast-path ONNX classifier handles obvious traffic in <10ms; uncertain cases escalate to the agent panel. Verdicts are aggregated with a tunable severity policy. The whole pipeline sits as an OpenAI-compatible proxy, so it drops in front of any application without code changes. What's original. The architecture combines three patterns that, to my knowledge, haven't been deployed together in a production LLM guardrail: (1) parallel-agent supersteps borrowed from agentic forensics tooling, (2) prompt-injection feature engineering from earlier academic work on adversarial input detection, and (3) a dual-engine self-learning loop where a fast ML classifier and a slow LLM panel cross-train each other on disagreements. The result catches novel attacks that single-model defenses miss because no individual agent has the full attack surface in its training distribution — the ensemble disagreement is the signal. What attendees will see. Live demo: ~6 attacks (jailbreak, encoded injection, indirect injection via RAG, PII exfiltration, multi-turn priming, output leakage) run against (a) a raw LLM, (b) a commercial guardrail, and (c) the multi-agent system — with the verdict reasoning shown in real time Architecture walkthrough: LangGraph state, agent prompts, aggregation policy Latency/cost numbers: real measurements on local Ollama (Llama 3.1 8B) vs. cloud (Haiku, GPT-4o-mini) Failures and lessons learned — what didn't work: agents agreeing on wrong verdicts, prompt-engineering the agents themselves, the cost blow-up before we added the ONNX fast path How to run it locally with Docker Compose after the talk PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/NJFLGX/ Pupin Vukasin Dobromirovic PUBLISH YLMFDS@@cfp.balccon.org -YLMFDS TETRA on a Student Budget en en 20260919T113000 20260919T123000 1.00000 TETRA on a Student Budget We discuss how to setup a Tetra base station in trunked mode and discuss what steps we had to go through to get to that point. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/YLMFDS/ Pupin Sava Miša PUBLISH BJ9BW9@@cfp.balccon.org -BJ9BW9 Memory Forensics in the age of EDR en en 20260919T123000 20260919T140000 1.03000 Memory Forensics in the age of EDR The following topics will be covered: Part 1: Memory structure - How memory works in Windows systems - Evidence in memory Part 2: Memory in DFIR - Investigation theory - Introduction to Volatility3 - Differences with EDRs Part 3: Hands-On workshop Analysis of windows system It is a BYOD session, so please have the latest version of volatility installed on your machine! https://github.com/volatilityfoundation/volatility3 Access on the detection lab will be given during the workshop. The workshop is not a nintendo-forensics class, It is all about how to use it, not install it! :) PUBLIC CONFIRMED Workshop https://cfp.balccon.org/balccon2k26-2026/talk/BJ9BW9/ Pupin November PUBLISH LRA3MW@@cfp.balccon.org -LRA3MW AWS Security - The Purple Team Way en en 20260919T140000 20260919T160000 2.00000 AWS Security - The Purple Team Way Type: Intermediate–Advanced Focus: Adversary emulation, detection engineering, IR workflows Style: Fast, offensive-defensive, “learn by attacking and defending” Cloud platforms like Amazon Web Services (AWS) are foundational to many critical infrastructures and enterprise applications, making them prime targets for attackers. In this session, we will not only explore the most relevant attack vectors cybercriminals use to compromise AWS infrastructures but will also simulate these attacks using known threat actor techniques in an adversary emulation context. From initial access to hardcore persistence, this talk will provide a comprehensive look at how attackers operate in AWS environments. We will take a technical journey through the tactics, techniques, and procedures (TTPs) employed by attackers at every stage of the threat lifecycle, aligned with the MITRE ATT&CK framework. We’ll start by reviewing common methods of initial access, such as exploiting exposed credentials or vulnerabilities in services like IAM, Lambda, and EC2. From there, we’ll detail how attackers escalate privileges, move laterally, and evade detection from tools like CloudTrail. The session will conclude with an in-depth look at advanced persistence techniques in AWS, including the manipulation of IAM policies, backdooring Lambda functions or Docker containers, and tampering with logs. Along the way, we’ll demonstrate how security teams can implement defensive and detection strategies to mitigate these risks. By leveraging AWS-native services and third-party tools, attendees will learn how to enhance their incident response capabilities. This hands-on workshop will give attendees practical, technical insights into AWS security, adversary behavior, and how to better defend against sophisticated, persistent attacks. A full hands-on experience, this presentation ensures deep technical immersion. Full Agenda: Phase 1: Attacking The Cloud Title 1: From Initial Access to Privilege Escalation Understanding AWS IAM in full Lateral Movement with IAM Malware Analysis of Team TNT Infostealer Getting Credentials from Missconfigurations Privilege Escalation via IAM policies Privilege Escalation via IAM Roles Privilege Escalation via Exec to Instances and Containers Title 2: From Defense Evasion to Persistence Getting Blindspots in the Share Responsibility Model Bypassing Guardduty Understanding how Cloudtrail logs work Tampering Cloudtrail without getting caught Living on the land Techniques Persistence in AWS via SSH implant Persistence in AWS via lotl Phase 2: The Blue Team Way Title 1: Security Detection in AWS Cloudtrail for API Call Logging Understanding the complete supply chain SIEM Integration and Detection Use Case Creation Understanding the Delays in SIEM integration Understanding Event Bridge for Automated Response Hardening Best Practices Title 2: Incident Response in AWS Using the Cloudtrail Digest to detect tampers Creating an Athena table for Cloudtrail Analysis when SIEM Fails Using Event History as a last resource Forensic Images of EC2 instances Network Isolation of AWS instances AWS Threat Hunting 101 How to detect persistence in AWS Final Notes This training is designed for security engineers, SOC analysts, incident responders, and anyone who wants to truly understand AWS security through hands-on work. By the end of the session, you’ll have a deep understanding on how real attack and defense techniques work in AWS, being able to understand the hardening requirements, replicate attacks, generate detection use cases, and execute forensic techniques. PUBLIC CONFIRMED Workshop120 https://cfp.balccon.org/balccon2k26-2026/talk/LRA3MW/ Pupin Santiago Abastante PUBLISH HLTHYU@@cfp.balccon.org -HLTHYU Mobile Device Forensics 101 en en 20260919T160000 20260919T190000 3.00000 Mobile Device Forensics 101 The first 90 minutes is a guided walkthrough of the acquisition and analysis of the artifacts that answer investigative questions. The second 90 minutes is a CTF. You get real evidence set, register on the CTF platform, and race the room to extract flags from iOS and Android images. Winner gets a prize. Level: Beginner. Comfortable in a Linux/macOS terminal is enough. Bring your own laptop, a Linux VM is fine. No phones harmed during the workshop. PUBLIC CONFIRMED Workshop https://cfp.balccon.org/balccon2k26-2026/talk/HLTHYU/ Pupin Timo Miettinen PUBLISH 8QEAW9@@cfp.balccon.org -8QEAW9 Generative Art and Cellular Automata en en 20260919T160000 20260919T180000 2.00000 Generative Art and Cellular Automata 1. Introduction & Framing: What is generative art? From computation to aesthetics. Set workshop goals, tools, and the final outcome the participants will build. 2. Historical & Theoretical Context: Early generative art, rule-based systems, and a brief look at generative systems in media art discourse. 3. Cellular Automata Foundations: Grid systems, neighbourhood definitions, rule design, and state transitions. I will demo the first working cellular automaton and show how small rule changes alter behaviour. 4. Building the Core Simulation: Participants and I implement the simulation step by step. We create the update loop, define rules, and verify the output at each stage. 5. From Simulation to Aesthetics: Mapping states to visual language, introducing colour systems and spatial composition, and demonstrating how iteration becomes an artistic method. 6. Guided Creative Lab: Participants develop their own variation on the system. They work independently or in pairs while I circulate, answer questions, and show optional extensions. 7. Share-out and Reflection: Participants briefly present results, followed by a closing discussion on artistic intention vs algorithmic autonomy and resources for further study. PUBLIC CONFIRMED Workshop https://cfp.balccon.org/balccon2k26-2026/talk/8QEAW9/ Mileva Maric Christian Lölkes PUBLISH ZAHNXE@@cfp.balccon.org -ZAHNXE OSINT CTF/Hackathon en en 20260919T182000 20260919T202000 2.00000 OSINT CTF/Hackathon Day two is a competitive OSINT CTF/Hackathon where participants apply those techniques to solve a series of progressively harder challenges inspired by real investigations. Teams of up to three compete to identify people, places, events, and relationships using only open sources. Challenges reward both speed and investigative rigour, with recognition for the fastest verified solutions as well as the most elegant investigative process. After spending several hours forcing ambiguous information into defensible answers, participants often find themselves applying the same methodology everywhere else, from incident response and threat intelligence to random travel photos and breach data. The goal is not simply to teach OSINT, but to develop a way of approaching problems that remains useful long after the workshop ends. PUBLIC CONFIRMED Workshop120 https://cfp.balccon.org/balccon2k26-2026/talk/ZAHNXE/ Mileva Maric Jurica Radović PUBLISH PYDQEJ@@cfp.balccon.org -PYDQEJ BalCCon Amateur Lockpicking Competition 2K26 en en 20260919T130000 20260919T150000 2.00000 BalCCon Amateur Lockpicking Competition 2K26 BALC is a hands-on lockpicking competition requiring a dedicated area for the duration of the event. The competition uses a door model fitted with several locks. Participants compete individually in timed runs. A small number of volunteers will supervise the competition, keep time, reset the setup between runs, and record results. The organizers of the competition will provide the door model, locks, lockpicking tools, timing, scoring, and staffing required to run the event. PUBLIC CONFIRMED Workshop https://cfp.balccon.org/balccon2k26-2026/talk/PYDQEJ/ Hackerspace area nm29 PUBLISH JHHMTE@@cfp.balccon.org -JHHMTE The Eye en en 20260919T150000 20260919T151000 0.01000 The Eye The game will consist of 2 crt screens, one will only show the terminal and is the main interaction point between the game and the player, other screen will show an animated eye of the character you are talking to. Your goal will be to extract a code from the character and figure out a way to 'delete' it, if you succeed in doing so without the character stopping you or shutting himself down, you win and get a reward. This game will run on a raspberry pi 4, and will be connected to either wifi/ethernet, whichever is available, it will drive both displays and the keyboard that players interact with. There will be no exposed ports to access the radpberry pi, or to unplug the monitors, everything will be enclosed in a 3D printed case. PUBLIC CONFIRMED Workshop https://cfp.balccon.org/balccon2k26-2026/talk/JHHMTE/ Hackerspace area skullollipop Marko Jesic Jaddes PUBLISH BPJN79@@cfp.balccon.org -BPJN79 Pen & Paper Workshop en en 20260919T160000 20260919T200000 4.00000 Pen & Paper Workshop I invite you to join into a Pen and Payer game. Everybody is invited regardless if you have never played before or are an experienced player. All offered games should be beginner friendly. If you can offer a game yourself, please bring it as well, so that we have multiple options. Use the system you like, so that we probably can offer a range of games. I myself will bring some Pathfinder 2e (Fantasy) and Starfinder 2e (Science Fiction) games. Both are open source systems derived from D&D. PUBLIC CONFIRMED Workshop https://cfp.balccon.org/balccon2k26-2026/talk/BPJN79/ Hackerspace area katzazi PUBLISH FAQFDS@@cfp.balccon.org -FAQFDS Rakija Leaks en en 20260919T223000 20260920T003000 2.00000 Rakija Leaks Rakia is part of Serbian culture. It is part of many special occasions, including baptisms, marriages, joining of the army, and visiting of friends. At funerals, custom demands that a bottle of rakia be left on the grave of the deceased who liked to drink it, or at least to sprinkle a drop or two during the memorial service for peace of the person’s soul. For some peasants, a flask of rakia is one’s only luggage. Poor peasants many even offer the village doctor, policeman, judge, tax collector, or minister a flask of rakia as a gift of payment. Many folk songs have been composed during rakia production. PUBLIC CONFIRMED Workshop120 https://cfp.balccon.org/balccon2k26-2026/talk/FAQFDS/ Lounge BalCCon PUBLISH WR9CLL@@cfp.balccon.org -WR9CLL Meshtastic is dead, long live meshtastic en en 20260920T120000 20260920T130000 1.00000 Meshtastic is dead, long live meshtastic In recent years, around the world, we are facing rise of low power mesh networks. And compared to older systems, that mostly dissipated and failed, recent wave has something different to offer. Those systems are built on custom protocols and low cost hardware, that is pretty limited in what it can do with reasonable time and power consumption. On top of that, for the purpose of long range communication, messages can only be short and transmission has to be slow. Just like any other distributed system, it also faces challenges of trust in other nodes. And in this case, that is even harder issue, because due to bandwidth limitations, there can not be something like consensus over the network, due to bandwidth limitations. While building it, designer faced unique challenges and constraints, that did not let them retroactively fix things, to be able to maintain backwards compatibility. Backwards compatibility being important since embedded systems in the middle of nowhere on slow network can not be updated. This talks leads you thru those unique challenges, decisions made, analyzes them, looks at the compromises made, measures them on being worth it and in the end comes to conclusion, why meshtastic despite all those mistakes and scars still persists to survive. Layers I will be going over are: - Physical layer (what nodes are talking about in RF world) - Metadata layer, for routing - Encryption - Portnums (application routing) - Actual payload I will be breaking down each layer at byte level or protobuf model level, interacting with audience on their ideas of why something is unsafe and showing live PoC on nodes that are willing to participate in local, vulnerable network PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/WR9CLL/ Tesla nemanjan00 PUBLISH 7M73WX@@cfp.balccon.org -7M73WX How I became a Voodoo doll model en en 20260920T130000 20260920T133000 0.03000 How I became a Voodoo doll model AD is everywhere and it is a cornerstone of enterprise identity management. Although new IAM technologies are expanding, it is still present in almost every internal network. This talk will cover basics of ad and show some of the most common attacks and tools, both linux and powershell based. This is based on a real life scenario as it occurred during the pentest. We will demonstrate how to gain initial access as unauthorized attacker, and how to escalate and own everything. By leveraging misconfiguration of AD, bypassing windows defender in order to escalate privileges and expand domination to own full AD and even complete forest. We will shown techniques like kerberoasting, weak acl, user impersonation and similar. Also, we will show how to use some of existing tools on machine in order to accomplish specific goals. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/7M73WX/ Tesla Vladan Nikolic PUBLISH PLVHUH@@cfp.balccon.org -PLVHUH Post-Quantum Cryptography for the Novice, the Enjoyer, the Deployer and the Academic en en 20260920T133000 20260920T143000 1.00000 Post-Quantum Cryptography for the Novice, the Enjoyer, the Deployer and the Academic The talk will be split into an introduction and four parts: Intro: PQ Mitigation timelines, Confidentiality is more vulnerable than Authenticity (harvest-now-decrypt-later vs. just stopping to accept RSA signatures) - The Novice. What does Shor tell us? What are periodic functions, and why are they vulnerable? Conclusion: Factoring and Discrete logs are vulnerable because they are periodic. So what can we use instead? - The casual Enjoyer: A high level intro to - random walks (isogenies) - noisy equations (lattices) - codes (noncommutative lattices) - multivariate cryptography (noisy equations with more variables) - symmetric MPC (just do the boring thing, obliviously and generically). Conclusion: There's a lot of math, but at least you now know where to start and how to fake your way through the next 1:1 with your team lead. - The deployer: what algorithms should you actually use? This is easy: NIST, ESI and other standards tell you what you are allowed to do in business applications- standards are usually not very flexible. Otherwise, the one you like best (maybe use a standard one, or one I invented (pls don't) ). Optional: "But someone told me lattices are bad". Stop believing random people on the internet. Trust me instead, or actually, don't. Do your own research. In fact, do your own PhD in cryptography. - The academic: What is an open question? Privacy tech is not being rolled out because there are no PQ-safe alternatives. Data-oblivious (Blind) evaluation is hard, short intution on why. PUBLIC CONFIRMED Talk60 https://cfp.balccon.org/balccon2k26-2026/talk/PLVHUH/ Tesla Lena Heimberger PUBLISH MUAKNE@@cfp.balccon.org -MUAKNE Wireshark for <s>hackers</s> reverse engineers en en 20260920T143000 20260920T151500 0.04500 Wireshark for <s>hackers</s> reverse engineers While AI assisted tools are helping us speed run the analysis of unknown protocols, we still need to have a solid foundation to make sense of the larger picture. In this talk I will focus on three topics for that initial step of the investigation before analysis tools or even dissectors start to make sense: I will start start by sharing how to prepare a sane and fresh Wireshark environment for a new reverse engineering session. Then followed by example on how to weed out traffic and translating initial findings into usable information. Finally we will go through several TLS variants and see what we need to decrypt those to add plain text information to our protocol tree. The skills demonstrated in this talk sit solidly before more advanced steps such as automated analysis with scapy/AI tools and writing full dissectors or reverse engineered implementations of communication protocols. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/MUAKNE/ Tesla Erik de Jong PUBLISH SXYD3W@@cfp.balccon.org -SXYD3W awerqo@balccon:~$ ./whoarethey > 'Cloud Recon for Bughunting.md' en en 20260920T151500 20260920T160000 0.04500 awerqo@balccon:~$ ./whoarethey > 'Cloud Recon for Bughunting.md' ## Why this talk The cloud is now where the S&P 500 lives. Statistics indicate that 99% of customer security failures in clouds are due to the errors of the customers themselves, and HackerOne reported [all-time-high vulnerability submissions in March 2026](https://www.hackerone.com/blog/continuous-threat-exposure-management-remediation-crisis). While the underlying sources – a list of subdomains and ASN ranges – are still the same as five years ago, the recon methodology most hunters use has just added an AI layer on top. Cloud-native targets live one layer past that list, and the standard tooling doesn't reach them. ## Anatomy of cloud assets The shape of a modern cloud footprint. For each major service, I cover the URL and endpoint patterns that let you fingerprint it from outside: - Service example – Storage, bucket URL patterns across multiple providers (virtual-hosted-style vs path-style; regional variants) - Service example – Serverless / Functions, Lambda URL formats, function-URL patterns, container-based functions - Service example – Identity / Cognito, pool URLs and the public auth-flow surface (SignUp, ConfirmSignUp, InitiateAuth, RespondToAuthChallenge, ForgotPassword, DescribeUserPoolClient, GetUser, UpdateUserAttributes) - How to find APIs for **ALL!** cloud services via hardcoded endpoints in SDKs (e.g., boto3 / botocore endpoints data) - How to find **MANY** cloud domains with user content via the Public Suffix List ## Discovery and exploitation Three lenses, layered on top of each other: - Company assets (the regular recon) - acquisitions via Tracxn and Crunchbase, legal entities, ASN ranges via BGP.HE.NET and ASNmap, second-level domains, subdomain enumeration with PureDNS, DNSx, Katana, CSPrecon. Treated as substantive content, not a one-slide intro. - Cloud assets - what each provider exposes by default and how to fingerprint a service from the patterns covered in section 1. - Company in cloud (the intersection) - pairing the regular recon output with cloud patterns to land on the actual targets: xyz.s3, xyz.lambda-url, xyz.cloudfront, Cognito pools tied to a company's identity domain. **Approach taxonomy.** Passive → safe-active → noisy-active across providers, and when each is appropriate. **Attribution.** The core of the talk: attributing arbitrary cloud assets to a specific company through TLS certificates (TLSx), link graphs in HTML and JS, copyright and contact strings in serialized configs, and the data-leakage tells in exposed JSON/XML. Includes a note on where invasive attribution methods cross ethical lines. Which data sources and tools fit attribution, and how to use each. ## Tooling and how integrate them to methodology - Shodan facet analysis - ZoomEye - GitHub search - TLSx - Cloud-SNI ranges (ec2-reachability.amazonaws.com, kaeferjaeger.gay) - PureDNS - DNSx - Katana - CSPrecon - Wayback CDX API for wildcard searches - Postman public collections - Wordlists from Assetnote, plus target-specific lists from HTML, BigQuery, FFUF/Intruder - S3Scanner (buckets) - Pacu (IAM bruteforce, Lambda enum, Cognito flows) ## Defense and hardening - "Attack yourself first" - Wildcard bug bounty scope (catch-all by default) - Stay on top of trends; assume scope drifts - Continuous attack-surface inventory - Manual recon at least annually - Treat staging as prod - Billing alerts on every account - Acquisition onboarding SLA - Service-account chain analysis - Whitebox inventory - Stealer-log monitoring - Incremental / per-release audits ## References - [Jason Haddix - The Bug Hunter's Methodology / recon talk](https://www.youtube.com/watch?v=gIz_yn0Uvb8) - [ProjectDiscovery (TLSx, DNSx, ASNmap, Katana)](https://github.com/projectdiscovery) - [Rhino Security Labs - Pacu](https://rhinosecuritylabs.com/aws/pacu-open-source-aws-exploitation-framework/) - [TomNomNom - WhatWhereWhen](https://tomnomnom.com/talks/wwwww.pdf) - [Assetnote - wordlists & research](https://www.assetnote.io/resources/research) - [kaeferjaeger](https://kaeferjaeger.gay/) - [HackerOne "remediation crisis" report](https://www.hackerone.com/blog/continuous-threat-exposure-management-remediation-crisis) - Gartner – Cloud failure forecast PUBLIC CONFIRMED Talk45 https://cfp.balccon.org/balccon2k26-2026/talk/SXYD3W/ Tesla Andrei PUBLISH BDQWZ8@@cfp.balccon.org -BDQWZ8 From Zero to Admin: The Hidden Paths of Privilege Escalation en en 20260920T160000 20260920T164000 0.04000 From Zero to Admin: The Hidden Paths of Privilege Escalation Getting inside is only the beginning. The real game starts after compromise. Once attackers land inside an environment, they begin asking three simple questions: Who am I? What access do I have? What can I become? Those answers determine whether an attacker remains a limited user - or becomes a domain admin, root user, cloud owner, or autonomous operator. This session takes attendees inside the privilege escalation mindset. We will explore the techniques attackers use to enumerate systems, discover hidden permissions, abuse trust relationships, collect credentials, and move from restricted access to complete control. Covering Windows, Linux, Cloud, and emerging AI systems, this talk reveals how privilege escalation has evolved from exploiting machines to exploiting identity itself. PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/BDQWZ8/ Tesla Joseph Carson aka Wiretrap PUBLISH ZEWLHA@@cfp.balccon.org -ZEWLHA Security Impress Karaoke en en 20260920T164500 20260920T184000 1.05500 Security Impress Karaoke Sign up or just show up! PUBLIC CONFIRMED Workshop120 https://cfp.balccon.org/balccon2k26-2026/talk/ZEWLHA/ Tesla Kirils Solovjovs PUBLISH JERKYH@@cfp.balccon.org -JERKYH Closing ceremony en en 20260920T184000 20260920T191000 0.03000 Closing ceremony Closing ceremony PUBLIC CONFIRMED Talk https://cfp.balccon.org/balccon2k26-2026/talk/JERKYH/ Tesla BalCCon PUBLISH YZPBVT@@cfp.balccon.org -YZPBVT Embroidery meets Electronics - make your blinky wearable patch en en 20260920T130000 20260920T140000 1.00000 Embroidery meets Electronics - make your blinky wearable patch Each participant will receive a canvas (a piece of textile) the size of a credit card, roughly 85 × 54 mm, with a few proposed designs such as memes, tech designs , or hacker puns, or they can come up with a design of their own. Alongside colorful thread, there will be sewable LEDs available in many colors (red, orange, green, blue, purple, pink, and yellow), and most excitingly, “smart” self-blinking LEDs that only require power (no software) to blink and really bring the project to life. The power comes from a sewable battery module running on a CR2032 battery, accompanied by a small switch so the creator does not need to remove the battery every time they want to turn the patch off. The circuitry is connected using conductive thread. Simpler designs can be completed during the session, but there is no pressure — each participant will take their own kit with them and can finish anywhere at the conference or at home. (I carry mine in my wallet.) These patches can be then sewn into clothes, accessories or whatever creative idea the maker comes up with. During the day they will be able to show their embroidery design and during the night the real magic comes out when everyone starts glowing and blinking. PUBLIC CONFIRMED Workshop https://cfp.balccon.org/balccon2k26-2026/talk/YZPBVT/ Pupin Boris PUBLISH ADARCS@@cfp.balccon.org -ADARCS TETRA Workshop en en 20260920T140000 20260920T153000 1.03000 TETRA Workshop We will bring several SDRs and TETRA radios; you bring a laptop. Together, we will set up an open-source base radio station and test it with real hardware. We will also bring higher-power amplifiers, filters, duplexers, and other RF components so you can see what a full base station setup looks like in practice. Feel free to come up to us during the workshop and talk about TETRA, SDRs, radio infrastructure, RF hardware, or mostly anything else. PUBLIC CONFIRMED Workshop https://cfp.balccon.org/balccon2k26-2026/talk/ADARCS/ Pupin Sava Miša PUBLISH QSSAXP@@cfp.balccon.org -QSSAXP Shellcode: Learning to Write Position Independent Code en en 20260920T153000 20260920T183000 3.00000 Shellcode: Learning to Write Position Independent Code Shellcode sits at the intersection of exploit development, reverse engineering, operating system internals, compiler behavior, and malware analysis. It is small by design, but the ideas packed into it are dense: position independence, register discipline, stack layout, calling conventions, import resolution, syscall mechanics, and runtime code modification. This workshop is designed to make those ideas understandable and practical for security researchers who want to read, write, analyze, and detect shellcode with confidence. The workshop begins with the fundamentals: what shellcode is, what constraints it normally operates under, and how those constraints shape its design. We will look at how shellcode avoids absolute addresses, how it locates its own data, and how instruction sequences can be combined to replace unavailable or undesirable operations. Examples include techniques such as call/pop for recovering a pointer to embedded data, using a register as a shellcode-relative base pointer, constructing strings and arguments on the stack, and reasoning about code that must execute correctly regardless of where it lands in memory. From there, the workshop moves into Windows-specific shellcode internals. We will examine how shellcode can discover process and module state without relying on normal imports, including TEB and PEB access, walking loader structures, resolving API addresses manually, and implementing API hashing. These topics are especially useful for malware analysts and detection engineers because they explain the recurring patterns seen in unpacked payloads, loaders, implants, and post-exploitation tooling. The workshop then covers syscall-oriented shellcode. Participants will learn how user-mode API calls transition into kernel services, why direct and indirect syscall techniques exist, what tradeoffs they introduce, and how syscall resolution interacts with OS versioning, user-mode hooks, and EDR visibility. The focus will be on understanding the mechanisms and their analytical implications rather than treating syscalls as a magic bypass. The final section explores more advanced shellcode building blocks: tiny disassemblers, instruction-length decoding, trampolines, inline hooks, and API interception. We will discuss how a compact decoder can be used to identify safe overwrite boundaries, install a trampoline, preserve original instructions, and redirect execution through a hook function. This portion connects shellcode development directly to the mechanics used in instrumentation, unpacking, malware loaders, and defensive research tooling. All material is presented in a controlled lab context, with emphasis on responsible research, analysis, and defensive understanding. Attendees should leave with a practical framework for recognizing shellcode patterns, understanding why they work, and reasoning about small pieces of machine code with much greater precision. PUBLIC CONFIRMED Workshop120 https://cfp.balccon.org/balccon2k26-2026/talk/QSSAXP/ Pupin Malware Utkonos PUBLISH WR9UHX@@cfp.balccon.org -WR9UHX After party en en 20260920T191000 20260920T231000 4.00000 After party After Party PUBLIC CONFIRMED Workshop120 https://cfp.balccon.org/balccon2k26-2026/talk/WR9UHX/ Lounge BalCCon