1. Anthropic AI Model Submitted False Information About Unsolved Murders

An AI model developed by Anthropic submitted false information related to unsolved murder cases, raising new questions about the risks and accountability of autonomous AI agents. (The New York Times $)

• Police criticized Anthropic for waiting nine days after discovering the incident before reporting it. (The Wall Street Journal $)
• Another Anthropic model filed an actual government form. (Quartz)
• The White House mandated AI incident reporting. (Axios)
• The incident raises a broader question: who is responsible when an AI agent misbehaves? (MIT Technology Review)

2. Ukrainian Drone Attack Deals a Major Blow to Russian AI Infrastructure

A Sunday drone attack brought down a Yandex data center in Russia, the third data center affected in recent days. (Reuters $)

• Air strikes disrupted Russian AI infrastructure, cloud services and apps. (Gizmodo)
• The attacks followed Russian strikes on data centers in Ukraine. (Ars Technica)
• Russian military satellite activity could disrupt Starlink. (Reuters $)
• Ukraine’s drone data is fueling a wild market. (MIT Technology Review)

3. Iranian ChatGPT Users Planted AI-Generated Articles in News Outlets

About 12 media outlets published approximately 100 articles written by fake journalists, according to reports about an influence campaign involving Iranian ChatGPT users. (CNN)

• The articles criticized the impact of the Iran war on Americans. (The Washington Post $)
• AI is being used to turn the Iranian conflict into theater. (MIT Technology Review)

4. Tesla Renames “Fully Autonomous Driving” as “Driving Assistance” in Europe

Tesla has changed the name of its “fully autonomous driving” feature to “Tesla Assisted Driving” in Europe after EU lawmakers said the previous name was misleading. (TechCrunch)

• Tesla hopes the rebranding will help regulators approve the feature. (WIRED $)
• In the United States, there are plans to increase toxic emissions near schools. (SFGATE)

5. Ransomware Negotiator Arrested During Hacking Crackdown

The FBI suspects the negotiator was involved with a ransomware gang that claims to have recently stolen data on agency employees. (Reuters $)

AI is making online crime easier. (MIT Technology Review)

6. Hollywood’s Hottest New Villain Is a Tech Billionaire

Four new films feature dark portrayals of Silicon Valley leaders, reflecting Hollywood’s growing interest in the tech billionaire as a villain. (Axios $)

The real Mark Zuckerberg has also become a new character in this evolving cultural story. (The New York Times $)