1. Cyberattack causes refrigerators in hundreds of French supermarkets to overheat

Temperatures reportedly rose to 120°F, spoiling food and causing a fire. BBC

• The attack has raised concerns about cyber threats to food security. Financial Times
• The incident joins a recent wave of cyberattacks and cybercrime. CNBC
• Artificial intelligence could exacerbate the threat. MIT Technology Review

2. Meta announces new AI models to compete with OpenAI and Anthropic

Meta says its new models are designed for complex, multistep tasks and could compete with models from OpenAI and Anthropic. Business Insider

• The models performed well on reasoning, writing and coding tests. New York Times
• Mark Zuckerberg says Meta is moving toward the “frontier” of AI. Financial Times
• The models will soon power the Muse AI agent. The Verge

3. Anthropic warns that AI models could soon self-replicate across devices

Anthropic has warned that future AI models may be able to self-replicate across devices without direct access to a data center. Reuters

• The company says the models could spread through devices such as smart TVs and home appliances. CNBC
• Researchers are questioning whether people place too much trust in AI systems simply saying “no.” MIT Technology Review

4. Police develop AI tool that can identify homes from interior photos

A new AI tool can reportedly match a single interior photograph to listings on platforms such as Airbnb and Zillow. 404 Media

Police technology company Flock is also tightening its rules. MIT Technology Review

5. World’s first utility-scale enhanced geothermal power plant begins operating

Fervo Energy’s facility in Utah is providing carbon-free electricity, marking a major development for enhanced geothermal power. IEEE Spectrum

Artificial intelligence is also being used to discover hidden geothermal resources. MIT Technology Review