Article · The Verge · Saturday 7 November

Why a new report on Anthropic automating AI development is alarming researchers

Anthropic is using AI to build more powerful AI. And it appears to be working remarkably well, with the company's models improving roughly twice as quickly as a year ago, with little indication of slowing down.

Read at theverge.com The opening of the article at theverge.com.

In this story

ProviderAnthropic

A tool or model opens in the Heidelberg AI Curriculum: what it does, what it costs, how to start. A repo opens on GitHub.

AI models can often spot problematic behavior by other AI agents in logs, according to a study by Anthropic researcher Andy Hall. GPT-5 found about 98 percent of reward-hacking cases, but the more exhaustively the search is tuned, the more false alarms pile up. That's not yet enough for reliable safety checks.

ProviderAnthropic

2 h ago

VideoBart Slodyczka

The 7 Levels of AI Agents (And Which One You Should Be On)

OpenAI recently gave a talk called What Comes After Loops. Around the same time, Anthropic published research on what happens when you put lots of AI agents together. Two rival companies, and they landed in pretty much the same place: using one agent well is basically solved.

ProviderAnthropic

8 h ago

ArticleLatent Space

[AINews] Anthropic IPO: Whither Now?

November 6, 2026 — ARC Prize tests Inkling 2 Small, and Thinking Machines explains how it caught early checkpoints bypassing guardrails.

16 h ago

Why a new report on Anthropic automating AI development is alarming researchers

In this story

Related stories

Anthropic pushes ahead with pre-Thanksgiving IPO despite tech selloff, accepts lower valuation

AI auditors catch most agent misbehavior but flood reviewers with false alarms, new study finds

The 7 Levels of AI Agents (And Which One You Should Be On)

[AINews] Anthropic IPO: Whither Now?