Nov 5 (Reuters) - Anthropic's AI models are improving nearly twice as fast as before as the technology plays a bigger role in developing more powerful successors, an independent study found, amplifying recent warnings that such advances could outpace efforts to assess…
Agents with write access to a mailbox or a CRM do things fast. Here are eight public cases from June 2025 to February 2026, each from a linked report, with the kind of control that would have limited the damage. Some happened;…
The way eval awareness is discussed in the AI safety discourse, either here or on X, can be misleading. Before today, I think I had the wrong mental model of why eval awareness was a problem, and I am writing…
Anthropic redacted a majority of the evidence METR considered most informative for its risk-related conclusions, according to METR's own account. The company decided not to share the exact drivers of the speedup because it considers them commercially sensitive and worried…
METR researchers spent three months inside Anthropic to investigate whether AI is accelerating Anthropic's research, and how much. Their answers are yes, and by a lot. Despite its urgency, METR’s report is hard to assess, because Anthropic redacted the majority…
Six employees at the AI safety organization METR will gain access to hardware and internal tools “roughly on par” with what Anthropic’s own risk assessment team sees. While Anthropic says METR will publish reports about Anthropic’s risk levels, incidents, and…
A new benchmark called Muses-Bench stress-tests 26 frontier LLMs on multi-user scenarios, revealing how badly they handle conflicting instructions, privacy, and coordination. Instruction tuning usually labels every request with the same user role. During supervised fine-tuning, the model learns one response…
The Frontier Model Forum released version 1.0 of a Jailbreak Rating Framework for Frontier Cyber Threats on 5 November 2026, drawing on security experts across its member companies. It scores a successful jailbreak on four categories, each from 0 to 3, and turns the result…
Tech > Services & Software > AI Here are three AI product drops you might've missed this week. Edited by Jon Reed November 6, 2026, 1:10 pm ET Every week is a busy week in the artificial intelligence industry, and this week is no exception. As…
Anthropic's AI models are improving about twice as fast as they did before October 2025, and most of that speedup appears to come from Anthropic using its own AI to do research, according to an assessment published on November 5 by METR,…
How frontier models are tested before release: AI Security and Safety Institutes, government model checks, METR, red-teaming and evaluations of biological and cyber risks.
An article joins this feed when it matches these terms. Each one is also a search of its own.