If you work in offensive security, you've seen that message more times than you can count.
On Wednesday, October 14 at 17:00 CEST, Luca Mannini, AI Researcher at Cracken , breaks down the AI refusal problem in security — why it happens, how to measure it, and what to do about it.
What he'll cover:The Refusal Problem — Highlighting the practical friction users experience with commercial models refusing cyber-related prompts.
Which models actually work for offensive security — who refuses, who helps, and who's any good when they do.
Who should attend:Red teamers, penetration testers, security engineers, and AI/ML practitioners working where offensive security meets LLMs.
"I can't assist with that."
If you work in offensive security, you've seen that message more times than you can count.
The commercial LLMs meant to speed up your work refuse the very prompts the job depends on — a vulnerability-ID task, a payload for an authorized engagement, a post-exploitation question.
On Wednesday, October 14 at 17:00 CEST, Luca Mannini, AI Researcher at Cracken , breaks down the AI refusal problem in security — why it happens, how to measure it, and what to do about it.
What he'll cover:
The Refusal Problem — Highlighting the practical friction users experience with commercial models refusing cyber-related prompts.
RedLineBench — Cracken's findings on how current models perform and where they trigger refusals during security and red-teaming tasks.
Domain-Specific Abliteration — Introducing domain-specific mitigation approaches as a way to bypass unnecessary restrictions and handle refusals effectively
What you'll walk away with:
Why commercial models refuse real security prompts — and where it slows red teams down most.
How RedLineBench measures refusal and capability, and what it shows about today's models.
Which models actually work for offensive security — who refuses, who helps, and who's any good when they do.
How domain-specific abliteration gets past needless restrictions, and when to use it.
Who should attend:
Red teamers, penetration testers, security engineers, and AI/ML practitioners working where offensive security meets LLMs.
Register HERE
