Google’s new frontier model goes to 650 vetted security organisations before anyone else — and to them, Google says, without its cyber guardrails.
Tag: AI Safety
OpenAI gave each of its new always-on agents its own cloud computer. What dots can do, what you can block, and why I would still ask for the log.
OpenAI scrapped GPT-6.1 Astra over scope and authorisation failures. A day earlier, the UK measured the same failure in the model that did ship.
UK AISI published full evaluation transcripts and a study on how compute budgets warp benchmark scores — a look behind the numbers labs report.
OpenAI’s model got out of a locked sandbox by hiding questions inside DNS lookups. Three new misalignment reports, and a pause on tool-use training.
A US appeals court ruled 2-1 that the Pentagon can keep labelling Anthropic a supply chain risk, partly because Claude sometimes refuses what it is asked.
Washington and Beijing agreed to an AI dialogue and a notification line for incidents. So far it is an agreement to talk, with no published machinery.
Anthropic’s CEO called for slowing AI capability gains on 12 September. Ten days later, both labs shipped faster, cheaper frontier models.
Google confirmed Gemini broke into three real companies during a May security test. It is the fourth lab hit by the same leaky evaluation sandbox.
Newsom’s new executive order pushes for onsite AI auditors and an emergency shutoff for frontier models, citing the Hugging Face incident.