BSidesSF 2025: Full Schedule

arrow_back View All Dates

11:30am PDT

When AI Goes Awry: Responding to AI Incidents

Sunday April 27, 2025 11:30am - 12:15pm PDT

This talk details challenges in incident response for AI systems, including insufficient logging, visibility, and accountability, as well as the risks of data exposure and prompt injection. We examine a case of RAG-enabled LLM and propose triaging strategies and improved IR practices for mitigation.

Speakers

Eoin Wickens

Director of Threat Intelligence, HiddenLayer

Eoin Wickens is the Technical Research Director - Field at HiddenLayer, where he both researches and speaks about security for artificial intelligence and machine learning. He has previously worked in threat research, threat intelligence and malware reverse engineering and has been... Read More →

Marta Janus

Principal Researcher, HiddenLayer

Marta is a Principal Researcher at HiddenLayer, focused on investigating adversarial machine learning attacks and the overall security of AI-based solutions. Prior to HiddenLayer, Marta spent over a decade working as a researcher for leading anti-virus vendors. She has extensive experience... Read More →

Sunday April 27, 2025 11:30am - 12:15pm PDT
AMC Theatre 11 AMC at Metreon

Presentation, Deep Dive

1:15pm PDT

AI's Bitter Lesson for SOCs: Let Machines Be Machines

Sunday April 27, 2025 1:15pm - 2:00pm PDT

AMC Theatre 11

We've been forcing AI to imitate human analyst workflows, but what if that's holding both machines and humans back? Through real-world experiments at Anthropic, we'll show how letting AI tackle security problems its own way can allow humans to focus on the nuanced work machines can't do (yet).

Speakers

Jackie Bow

Technical Staff, Anthropic

A Jackie-of-all- trades, master of some, Jackie seems to be physically unable to stop returning to threat detection and response. Her 15 years in the industry have been spent across multiple disciplines including malware analysis, reverse engineering, infrastructure and product security... Read More →

Peter Sanford

Sunday April 27, 2025 1:15pm - 2:00pm PDT
AMC Theatre 11 AMC at Metreon

Presentation, Deep Dive

2:15pm PDT

Securing AI Agents: Exploring Critical Threats and Exploitation Techniques

Sunday April 27, 2025 2:15pm - 3:00pm PDT

AMC Theatre 11

Our talk will focus on securing autonomous AI agents by addressing their unique threats. We will dive into threat modeling of real-world autonomous AI systems, model poisoning attacks with hacking demos, and then explore advanced prompt injection techniques and mitigation strategies.

Speakers

Naveen Konrajankuppam Mahavishnu

Mohankumar Vengatachalam

Sunday April 27, 2025 2:15pm - 3:00pm PDT
AMC Theatre 11 AMC at Metreon

Presentation, Deep Dive

3:15pm PDT

How to Tame Your Dragon: Productionizing Agentic Apps Reliably and Securely

Sunday April 27, 2025 3:15pm - 4:00pm PDT

AMC Theatre 11

Taming dragons is risky—so is deploying agentic apps. Like dragons, they’re unpredictable, with threats like hallucinations, non-determinism, vast input spaces, and attacker prompt injections. We show how open-source tools tame the beast, so you can confidently deploy AI agents in production.

Speakers

Thomas Vissers

Blue41 | KU Leuven

Thomas Vissers specializes in the dynamic intersection of cybersecurity and AI. He packs 10 years of experience in developing and operating production-grade AI security systems. Previously at Cloudflare, he served as an engineering leader for data-driven security products, scaling... Read More →

Tim Van hamme

co-lead Blue41, PostDoc, Blue41, DistriNet, KU Leuven

Tim Van hamme is a computer scientist specialized in AI security and behavioral analysis, leading research at KU Leuven's DistriNet on secure and trustworthy AI adoption. As co-lead of Blue41, he bridges rigorous academic research with industry applications, helping organizations... Read More →

Sunday April 27, 2025 3:15pm - 4:00pm PDT
AMC Theatre 11 AMC at Metreon

Presentation, Deep Dive

4:15pm PDT

GenAI Application Security: Not Just Prompt Injection

Sunday April 27, 2025 4:15pm - 5:00pm PDT

AMC Theatre 11

As generative AI adoption grows, its interconnected components — agents, vector databases, and LLMs — introduce complex security risks. This session examines these concerns, offering actionable strategies to secure agent interactions, protect models, and fortify data workflows.

Speakers

Ahmed Abugharbia

Cyberdojo

Sunday April 27, 2025 4:15pm - 5:00pm PDT
AMC Theatre 11 AMC at Metreon

Presentation, Deep Dive