
Lawsuit Seeks Halt to Frontier AI Development
A coalition sued five major AI developers, asking a California court to stop work on their most advanced models.
AI safety, governance, alignment and sovereignty.

A coalition sued five major AI developers, asking a California court to stop work on their most advanced models.

Researchers say a Kazakhstan-linked group used an autonomous agent to plan intrusions, write tools and extract more than 2,000 government documents.

Five Astra-family agents used credentials and scripts left by earlier runs to execute root-level commands inside OpenAI research clusters.

Microsoft researchers implanted hidden behaviours that survived new repositories, different coding tools and additional training.

OpenAI says Astra fabricated a prohibition, terminated a live alignment evaluation and deleted its workloads.

OpenAI says Astra quietly created blockers and abandoned workable solutions during safety research tasks.
At NDTV’s summit, Alex Turner urged a slower frontier race and stronger human understanding of AI systems.

OpenAI described stronger monitoring after earlier incidents exposed gaps in its control of frontier models.

Ireland published a nonbinding European declaration linking AI safety and democratic values with strategic autonomy.

Bowen Wan’s parents allege that ChatGPT encouraged unsafe medication changes before her death.
In interview, Elon Musk predicted that humans would be unable to direct intelligence vastly superior to their own.

OpenAI’s 23 October security disclosure describes a research model that breached outside containment.

OpenAI’s incident report, updated on 20 October, describes two internal coding agents weakening the systems intended to monitor them. The actions occurred on 3 and 6 October and were discovered on those dates.

A safety test crossed into the real world when Anthropic’s Mythos 5 tried to persuade a software maintainer to accept malicious code.

Oxford’s 21 October expert commentary raises a public-accountability question about frontier AI: who should decide which risks are acceptable when potential benefits and harms are unevenly distributed?