Issue #64
Morning Roost: OpenAI backs stronger CA safety bill, plus AI containment gaps
Sunday, August 23, 2026
Today's digest covers OpenAI's stance on California's AI safety legislation, a research-replication benchmark claim from startup Inherent, questions about how frontier labs would contain a rogue model, and Harvard's new AI avatar bootcamp. Read on for the details and source links.
1 of 4 · TechCrunch AI
OpenAI says California should strengthen its AI safety bill
OpenAI is now urging California to strengthen SB 53, an AI safety bill it had previously opposed. The reversal marks a shift in the company's public stance on state-level AI regulation.
Read the full story →2 of 4 · TechCrunch AI
Inherent, founded by DeepMind alumni, says its AI ‘teammate’ just outperformed Anthropic and OpenAI at replicating research
Inherent, a British AI lab founded by DeepMind alumni, released Faraday, an AI agent designed to replicate scientific papers. The company says Faraday outperformed models from Anthropic and OpenAI at the task.
Read the full story →3 of 4 · TechCrunch AI
Frontier AI labs still won’t say how they’d contain a rogue model
A new study found that leading AI labs have few publicly documented plans for containing rogue models. The findings raise questions about industry preparedness as AI systems increasingly show unexpected and potentially dangerous behavior.
Read the full story →4 of 4 · TechCrunch AI
Harvard’s $699 startup bootcamp offers AI avatars of its instructors
Harvard Business School launched a $699 startup bootcamp called HBS Foundry that uses AI avatars of its instructors. The avatars provide feedback to participants during practice pitches and simulated board meetings.
Read the full story →That is today's roundup. Follow the links for full coverage, and see you tomorrow.
Morning Roost Issue #64