The Chertoff Group

Our blog today is a must-read on AI model misalignment and adversarial use of AI, and what this means for companies.
The piece includes insights on speed and scale, the potential for agents to magnify misalignment and insights for incident responders. Learnings from these developments are relevant to defending against AI-enabled attacks - every company has to deal with this and managing misalignment risk when deploying agents.
There's a link to our capabilities brief, "Defending at Machine Speed" at the end of the blog post and on our homepage. #AgenticAI #Cybersecurity
chertoffgroup.com/agentic-ai-sec…

This week we are thinking about 9/11 and how everything changed after that day. We're honored to be part of @VitalCityNYC's latest issue. vitalcitynyc.org/nyc-terrorist-…

In an interview this morning on @SquawkCNBC, Adam Isles tells host Becky Quick that while Iranian cyber trade craft is not new, the scale of the attacks on water systems in the US is bigger than we’ve seen previously. Resiliency and the ability to disconnect systems from the internet and keep them running are key. AI is lowering the knowledge gap for the bad guys, and the good guys have catch up work to do. cnbc.com/video/2026/08/…

It's an understatement to say #Mythos is a hot topic. We recorded this conversation before the USG's export control directive that led Anthropic to suspend access to its Mythos and Fable models. Our core themes hold, and the latest developments underscore how rapidly this landscape is shifting and why security leaders need to build defensible security programs. youtu.be/uXZLojZM37g

In this episode of The Holy Grail of Investing, Christopher Zook and Mark Wade sit down with former U.S. Secretary of Homeland Security Michael Chertoff and Co-Founder of The Chertoff Group, Chad Sweet, for a conversation on risk, global security, and the shifting investment landscape. youtube.com/watch?v=xnWMeJ…