"We're becoming completely paranoid!", "Friday it was panic, an AI entered almost all classified systems of the NSA in a few hours"
Twist in the case of the ban of Anthropic's AI Fable 5 and Mythos: the real reason behind it is finally revealed.
SAFETY
-
True reason for banning AI Fable and Mythos revealed after NSA breach
By
–
-
Mythos hacked NSA systems, export controls imposed on Mythos and Fable
By
–
Mythos broke into almost all of the NSA’s classified systems in hours, per its director. It would have been irresponsible to not impose export controls on it. (And on Fable, with its pathetically inadequate guardrails.)
-
Generative AI is a runaway train, says Gary Marcus
By
–
If Generative AI were a vehicle, it would be a runaway train.
-
Complete course on agentic AI with LangChain and LangGraph
By
–
One of the best courses on agentic AIs I have seen. Nearly 10 hours of excellent content. Covers LangChain, LangGraph, RAG, deepagents, guardrails, and more. Are there any other good Lang* resources for people interested in learning?
-
US government likely downloads AI models right after training
By
–
a related point is that every time an AI lab finishes training a model, the US government probably downloads it over the wire immediately the idea that a five-year old AI lab has fully secured itself from nation-state hacking is a bit naive https://
x.com/willdepue/stat
/willdepue/status/2068118253633737077
… -
Responsible innovation key to trustworthy AI at VivaTech 2026
By
–
Trustworthy AI begins with responsible innovation. The conversation at @VivaTech 2026 underscored the importance of developing AI systems that are safe, transparent and human-centric, ensuring that technological progress delivers meaningful impact while maintaining public trust.
-
Gary Marcus Lists Eight Reasons to Slow Hyperscaling
By
–
Eight reasons to slow down hyperscaling:
• What GenAI has already done to society
• What AI slop is doing to the internet
• What data center overbuild is likely to do the economy • What data center overbuild is likely to do the environment • What cyberattacks may do to the -
Stop AI sycophancy with critical paper review prompts
By
–
If your AI is still being sycophantic, here’s how to solve it Option 1: “here is my research paper – be critical of it please” Option 2: “here is my research paper – spin up subagents to review the paper from the lens of Statler and Waldorf and Siskel and Ebert and
-
US measures and Anthropic restrict access to AI
By
–
Over the past two weeks, both the US government and Anthropic have taken significant measures that demonstrated their power to control access to AI by restricting what others can do with cutting-edge models. This has been one of those
-
Research on anthropomorphic misalignment requires stronger evidence
By
–
« Position: Research on anthropomorphic misalignment needs stronger evidence » AI safety papers often claim that models are deceptive, calculating, or concerned with self-preservation, but many results do not show