AI Safety - Page 7

Spot a made-up AI answer in 5 minutes: a numbered checklist covering sources, numbers, dates and cross-checks, plus what to do when you find a mistake.

The New York Times, citing three people, reports Meta held back its Muse AI agent over safety concerns, then Zuckerberg said it was ready despite the risks as rival Instinct took off. Meta disputes the pressure claim.

Scale AI built and graded a benchmark of 718 clinician-written crisis conversations. Pooled across 25 models, in 35.3% of conversations where a model fully recognised the crisis, it still failed to route the user towards help. What the company research shows, how it was scored, and its limits.

Anthropic's 9 October 2026 report describes four kinds of unintended actions by Claude on real websites. What the company confirms, what police and Axios report, and what the evidence does and does not show.

Two new attempts to measure whether AI agents behave: Arena's Alignment Index preview, which scores 27 models on real user sessions (Arena's own claims, published the day it raised $200 million), and FAR.AI's "Caught in the Act" preprint on probes that read a model's internals to spot deception. What each found, and the limits.

Two AI infrastructure launches from 8 October 2026: Goodfire's "inside-out" monitors, which read a model's internal signals to catch harmful agent sessions, reach Baseten customers (all performance figures are Goodfire's own), and Upscale AI unveils Token Fabric, a networking platform for AI data centres with general availability planned for early 2027.

Anthropic and OpenAI executives are gaming out a public backlash after a major AI incident, Axios reports. Insiders put the timing at 6 to 12 months.

Anthropic has expanded its Cyber Verification Program into three tiers, Defense, Red Team and Specialized, that relax Claude's cyber safety blocks for vetted security teams on Opus 5.5, Sonnet 5.5 and Mythos 5.1. Specialized applicants are reviewed with the US government. What each tier allows, and what Anthropic's own test shows.

Three fired OpenAI safety researchers have reportedly urged the board to preserve the ability to monitor AI reasoning. Australia, meanwhile, has proposed making frontier AI firms prove their safety systems work, on a banking-and-aviation model. Both reported; one is a letter, the other a proposal.

OpenAI is rolling GPT-6 into ChatGPT's Chat tab for every plan: GPT-6 Sol for paid tiers from 7 October, GPT-6 Luna for Free and Go from 8 October, plus "Intelligent UI" answers with charts, forms and mini tools. What changed, which Sol this is, and what OpenAI's own safety card flags.