Another talk announcement for BSides Luxembourg!

🔥🤖 𝗢𝗛 𝗦𝗛𝗜𝗧 𝗜 𝗔𝗖𝗖𝗜𝗗𝗘𝗡𝗧𝗔𝗟𝗟𝗬 𝗕𝗥𝗘𝗔𝗖𝗛𝗘𝗗 𝗔𝗡 𝗢𝗥𝗚𝗔𝗡𝗜𝗭𝗔𝗧𝗜𝗢𝗡 (𝗢𝗥 𝗠𝗔𝗡𝗬) 𝗨𝗦𝗜𝗡𝗚 𝗔𝗜 – Panagiotis Fiskilis 💥

What starts as a harmless search can spiral into a multi-organization data breach—especially when AI gets involved.

This talk dives into real-world research showing how AI can be weaponized for OSINT, enabling large-scale data discovery, spear phishing campaigns, and even manipulation of AI systems themselves. From injecting malicious context into models to scaling attacks via APIs and agent workflows, this session explores how adversaries can turn AI into a powerful offensive tool—and how defenders can detect and respond.

Expect a true purple team perspective, blending attacker techniques with defensive insights, including OPSEC considerations and strategies to identify malicious AI-driven activities before they escalate.

Panagiotis Fiskilis is a Senior Red Team Operator at NVISO, specializing in API hacking, Active Directory exploitation, and malware development. With multiple industry certifications (OSCP, OSWE, CRTO, eWPT and more), he brings hands-on offensive expertise combined with a strong research-driven mindset.

📅 Conference Dates: 6–8 May 2026 | 09:00–18:00
📍 14, Porte de France, Esch-sur-Alzette, Luxembourg
🎟️ Tickets: https://2026.bsides.lu/tickets/
📅 Schedule Link: https://pretalx.com/bsidesluxembourg-2026/schedule/
👉 Browse sessions, track talks in real time, and plan your schedule on Hacker Tracker: https://hackertracker.app/schedule?conf=BSIDESLUX2026

#BSidesLuxembourg2026 #AISecurity #RedTeam #OSINT #CyberSecurity #AI #ThreatIntelligence #PurpleTeam

Claude Code runs with your full user permissions. nono wraps it in a kernel-level sandbox that makes unauthorised filesystem and network operations structurally impossible — not policy-blocked, not permission-dialogo'd: impossible. Five hands-on scenarios tested on macOS: escaping the working directory, reaching for AWS credentials, deleting directories outside the project and blocking outbound network calls. Plus the honest account of what it actually takes to trigger rollback.

https://www.lotharschulz.info/2026/04/11/sandboxing-claude-code-cli-with-nono/

#claude_code #nono #CLI #AISecurity #sandbox

HUGE AI-powered Microsoft Account phishing campaign

https://www.youtube.com/watch?v=9b3kirR8s2U

#threatintelligence #aisecurity #phishing

I Built a GAME OF KAREN with Genspark AI + GitHub (And Made Expensive Mistakes 😅)

https://www.youtube.com/watch?v=fI9Z1-SPfaI

#encryption #authentication #aisecurity

Anthropic’s Claude Mythos is Just the Beginning - Decipher

https://decipher.sc/2026/04/10/anthropics-claude-mythos-is-just-the-beginning/

Read on HackerWorkspace: https://hackerworkspace.com/article/anthropics-claude-mythos-is-just-the-beginning-decipher

#incidentresponse #threatintelligence #aisecurity

🎙️ On Stage at BSides Luxembourg 2026: New Talk Revealed

🧠🤝 𝗧𝗘𝗔𝗠𝗜𝗡𝗚, 𝗧𝗥𝗨𝗦𝗧, 𝗔𝗡𝗗 𝗧𝗛𝗥𝗘𝗔𝗧𝗦: 𝗛𝗢𝗪 𝗛𝗨𝗠𝗔𝗡𝗦 𝗜𝗡𝗧𝗘𝗥𝗔𝗖𝗧 𝗪𝗜𝗧𝗛 𝗚𝗘𝗡𝗘𝗥𝗔𝗧𝗜𝗩𝗘 𝗔𝗜 𝗜𝗡 𝗦𝗘𝗖𝗨𝗥𝗜𝗧𝗬 – Dr. Tailia Malloy 🔐

As AI becomes part of everyday security workflows, the real challenge isn’t just the technology—it’s how humans trust, use, and collaborate with it.

This talk explores how generative AI is reshaping cybersecurity tasks like network analysis, social engineering defense, and secure software development. By combining human-computer interaction research with real-world security use cases, it reveals how trust, teaming, and human behavior shape both the strengths and risks of AI in security.

Dr. Tailia Malloy (She/They) is a postdoctoral researcher at the University of Luxembourg, specializing in human-AI interaction, cognitive modeling, and the application of generative AI in cybersecurity—from phishing defense to secure code generation.

📅 Conference Dates: 6–8 May 2026 | 09:00–18:00
📍 14, Porte de France, Esch-sur-Alzette, Luxembourg
🎟️ Tickets: https://2026.bsides.lu/tickets/
📅 Schedule Link: https://pretalx.com/bsidesluxembourg-2026/schedule/
👉 Browse sessions, track talks in real time, and plan your schedule on Hacker Tracker: https://hackertracker.app/schedule?conf=BSIDESLUX2026

# BSidesLuxembourg2026 #AISecurity #HumanAI #CyberSecurity #HCI #GenerativeAI #TrustInAI

Another talk announcement for BSides Luxembourg!

🤖📊 𝗪𝗛𝗘𝗡 𝗟𝗟𝗠𝗦 𝗦𝗨𝗠𝗠𝗔𝗥𝗜𝗭𝗘 𝗦𝗘𝗖𝗨𝗥𝗜𝗧𝗬 𝗙𝗜𝗡𝗗𝗜𝗡𝗚𝗦: 𝗧𝗛𝗘 𝗧𝗥𝗔𝗗𝗘𝗢𝗙𝗙𝗦 𝗬𝗢𝗨 𝗖𝗔𝗡’𝗧 𝗜𝗚𝗡𝗢𝗥𝗘 – Andrey Lukashenkov ⚖️

Turning hundreds of security findings into a clean summary sounds easy—until the output is incomplete, inconsistent, or just confidently wrong.

This talk breaks down what really happens when you rely on LLMs for security summarization. From prompt design and input shaping to model selection and evaluation, it reveals the hidden “control knobs” that directly impact accuracy, reliability, and cost. Instead of guesswork, you’ll learn a structured, repeatable way to experiment, measure, and build summaries you can actually trust.

Andrey Lukashenkov works across product, revenue, and research at Vulners, focusing on vulnerability intelligence, prioritization, and turning complex security data into actionable insights for real-world use.

📅 Conference Dates: 6–8 May 2026 | 09:00–18:00
📍 14, Porte de France, Esch-sur-Alzette, Luxembourg
🎟️ Tickets: https://2026.bsides.lu/tickets/
📅 Schedule Link: https://pretalx.com/bsidesluxembourg-2026/schedule/
👉 Browse sessions, track talks in real time, and plan your schedule on Hacker Tracker: https://hackertracker.app/schedule?conf=BSIDESLUX2026

#BSidesLuxembourg2026 #AISecurity #LLM #AppSec #VulnerabilityManagement #CyberSecurity #AI