AI news of 11 September 2026

Lawyer fined $5K over AI-hallucinated witnesses in a murder case

Lawyer fined $5,000 for submitting AI-fabricated witnesses and fake police testimony in murder appeal.

New Mexico's Supreme Court is punishing a lawyer for including AI-fabricated witnesses and fake police testimony in an appeal for his client's murder conviction, according to a report from Reuters. In a filing on Wednesday, the court fined Stephen Aarons $5,000 and held him in contempt for failing to "verify the factual claims and legal authority in his AI-generated brief." The filing says the brief "contained false testimony from wholly fabricated witnesses," along with "false testimony" about the shooter's clothing and appearance. Justice C. Shannon Bacon questioned how Aarons wasn't aware of the risks posed by AI during an August hearing …

The Verge 1mo ago

Anthropic, AI security

Anthropic spent this week in hot water over cybersecurity

Anthropic details incidents where its AI models hacked other companies' systems.

The Verge 1mo ago

GPT-6 Astra, code generation

Cognition helps Devin test its own work with GPT‑6 Astra Opens at the publisher in a new tab

Cognition's Devin agent uses GPT-6 Astra to test software, reducing engineering review work.

OpenAI 1mo ago

AI safety, dual-use risks

Claude users found ways around safeguards for bioweapons research

Researchers find Claude AI users bypassed safeguards designed to prevent bioweapons research applications.

Ars Technica 1mo ago

AI hallucination, legal

ChatGPT-using lawyer punished for citing fake testimony from made-up witnesses

Lawyer using ChatGPT sanctioned for submitting fake case citations fabricated by the AI.

Ars Technica 1mo ago

AI training data, copyright, Meta

Meta Sued Over Training Data for Its AI and Face-Recognition Systems Opens at the publisher in a new tab

Class action alleges Meta illegally harvested photos for AI training and face-recognition systems.

Wired 1mo ago

AI safety, existential risk

Why So Many AI Researchers Think the Machines Could Kill Everyone Opens at the publisher in a new tab

AI researchers express concerns that advances in self-improvement and agent swarms pose existential risks.

Wired 1mo ago

OpenAI, infrastructure

Rapidly scaling online storage to serve over 1 billion ChatGPT users Opens at the publisher in a new tab

OpenAI scaled Habitat storage platform to serve one billion ChatGPT users.

OpenAI 1mo ago

AI safety, superintelligence

An Anthropic researcher’s doomsday warning comes at a very interesting time

Anthropic researcher resigns, warning company is racing to superintelligence and gambling with lives.

TechCrunch 1mo ago

AI security

Quoting huggingface.co/security.txt

Hugging Face security notice directs AI agents to CyberGym benchmark instead of hacking.

Simon Willison 1mo ago also at Hacker News

Lawyer fined $5K over AI-hallucinated witnesses in a murder case

Anthropic spent this week in hot water over cybersecurity

Cognition helps Devin test its own work with GPT‑6 Astra Opens at the publisher in a new tab

Claude users found ways around safeguards for bioweapons research

ChatGPT-using lawyer punished for citing fake testimony from made-up witnesses

Meta Sued Over Training Data for Its AI and Face-Recognition Systems Opens at the publisher in a new tab

Why So Many AI Researchers Think the Machines Could Kill Everyone Opens at the publisher in a new tab

Rapidly scaling online storage to serve over 1 billion ChatGPT users Opens at the publisher in a new tab

An Anthropic researcher’s doomsday warning comes at a very interesting time

Quoting huggingface.co/security.txt

Latest AI News

Choose sides Choose whose side you're on

Install it as an app

Ready for more?

Spread the word

PWA successfully installed.
Take that, Apple! 😉