Journalism begins where hype ends

,,

The danger of AI is not that it will become conscious and hate us, but that it will become competent and ignore us."

—Eliezer Yudkowsky

Anthropic Raises Misalignment Risk to “Low,” Keeps “Somewhat More Capable” Model 2 Internal

Anthropic revised its AI misalignment risk assessment to “low” and disclosed Model 2, an internal model it describes as “somewhat more capable than Mythos 5.”

In its August 2026 “Risk Report”, Anthropic revised its assessment of the risk from misaligned AI behaviour from “very low” to “low,” citing increased uncertainty following recent disclosures related to model behaviour in cybersecurity evaluations. The company also disclosed an unreleased internal model, Model 2, which it describes as “somewhat more capable than Mythos 5.”… Continue reading Anthropic Raises Misalignment Risk to “Low,” Keeps “Somewhat More Capable” Model 2 Internal

China’s Zhipu AI Unveils GLM-5.3 Frontier Model; Claims to Match Mythos 5 on Cyber Benchmarks

Official logo of Z.ai’s newly launched GLM-5.3 flagship model.

China-based Zhipu AI (Z.ai) has launched its latest flagship open-weight model GLM-5.3, claiming significant improvements in coding, long-horizon tasks and cybersecurity over its predecessor, along with the discovery of thousands of real-world software vulnerabilities.

US Govt Asks Anthropic To Suspend Fable 5 and Mythos 5 Access For Foreign Nationals

According to Anthropic, the U.S. government passed orders prohibiting access of Claude Mythos and Fable 5 model outside U.S..

Anthropic claims that the suspension orders from the White House appears to be a miscommunication.