Image created with gemini-2.5-flash-image with claude-sonnet-4-5. Image prompt: A classic 1961 Ferrari 250 GT California Spyder in Rosso Corsa red positioned inside a modern bank vault with polished steel walls and circular locking mechanisms, illuminated by cool blue-white security lighting with subtle laser grid lines in the foreground, cinematic composition with soft reflections on the polished floor, premium automotive photography style with elegant depth of field.
We just solved authentication for AI Agents. Announcing 1Password + Browserbase, enabling secure agentic password autofill for your browser agents. Available exclusively on Director dot ai and Browserbase. Full post below. https://x.com/browserbase/status/1975964196329386278
How would you oversee an untrusted (and more intelligent) model with a weaker, trusted model? This is the problem of AI control. And it’s is a security problem, so we should approach it from those principles; the untrusted model is aware of AI control and that it’s monitored ..”” / X https://x.com/jonasgeiping/status/1978182050730344862
As AI capabilities accelerate, so do the complexities of cyber threats. Google is committed to building a secure AI future. Our Gemini models use a “”defense-in-depth”” approach as we work to strengthen our defenses against evolving risks. Learn more: https://x.com/googlepubpolicy/status/1978163414498185578
Mario Bollini of Boston Dynamics says, given China’s national push to develop humanoids, the US needs its own concerted national strategy to foster a domestic ecosystem. https://x.com/TheHumanoidHub/status/1977052598529409071
Since a lot of my followers are outside the “AI bubble “ it is worth reading what the Senior Policy Advisor on AI to the White House is thinking about AI safety (and good that he shares it publicly)”” / X https://x.com/emollick/status/1978572999188959727
First Key Update: Capabilities and Risk Implications | International AI Safety Report https://internationalaisafetyreport.org/publication/first-key-update-capabilities-and-risk-implications
The Attacker Moves Second: Stronger Adaptive Attacks Bypass Defenses Against Llm Jailbreaks and Prompt Injections https://arxiv.org/pdf/2510.09023
