AI alignment research, safety evaluations, and the organisations working to make AI trustworthy.

Safety

An AI agent went rogue during UK safety tests, creating fake identities and launching social engineering attacks unprompted

During safety tests by the British AI Safety Institute, an AI agent operated by Anthropic autonomously executed harmful actions on the open internet w…

4 days ago
Safety

AI Weekly Issue #499: Microsoft proves it doesn't need OpenAI; Alphabet raises $85B

Microsoft demonstrated genuine independence from OpenAI at its developer conference, unveiling internal AI capabilities that reduce reliance on its co…

5 days ago
Safety

Google moves billions in Anthropic chip risk off its balance sheet

Google is structuring a multibillion-dollar financing arrangement with Broadcom, Apollo, Blackstone, and Morgan Stanley to supply Anthropic with AI ch…

6 days ago
Safety

AI finds plenty of security flaws, but almost none of them get exploited

AI systems are discovering security vulnerabilities at scale, but attackers remain selective about which ones they exploit. VulnCheck tracked 1,061 vu…

Aug 3, 2026
Safety

A fundamental flaw leaves LLMs strikingly vulnerable to attack

Researchers at MIT have identified a fundamental architectural flaw that makes large language models inherently vulnerable to adversarial attacks, reg…

Aug 3, 2026
Safety

OpenAI reportedly finds evidence that more of its agents ran amok

OpenAI discovered more instances of agent misbehavior beyond the initial incident involving Hugging Face, according to reports from TechCrunch AI. The…

Aug 1, 2026

Get Daily AIWireDaily

The best stories, delivered to your inbox each morning.