AI Governance | Tech Brew

Anthropic’s new fund comes with fine print

Two days after its internal data showed junior hiring shrinking, Anthropic launched a fund it says could help workers impacted by AI.

Lindsey Choo

On/Off switch turned to "off" on a pixellated background

The scramble for an AI off switch

Rogue AI agent incidents are piling up. Lawmakers want a kill switch, but Nvidia’s betting the fix is a product, not a policy.

Lindsey Choo

Photo illustration showing a burglar wearing a black balaclava tiptoeing away, holding a pixellated garbage back of stolen data. His black shirt has the Microsoft and OpenAI logos on it.

AI scraping is theft, admits a Microsoft exec

A senior Microsoft exec called the AI industry’s mass scraping “an astonishing theft”—and it could lead to more lawsuits.

Lindsey Choo

It’s just you, ChatGPT, and your human reviewer

OpenAI uses third-party human contractors to review user ChatGPT logs—ostensibly to improve its AI models.

Whizy Kim

Dario Amodei, Sam Altman, Elon Musk on a glitchy background

To pace or not to pace

Major US frontier labs mostly agree with Dario Amodei’s call for an AI slowdown. Unfortunately, the government doesn’t seem to be on the same page.

Whizy Kim

Animated gif from the Terminator movie, showing a humanoid robot with glowing red eyes holding a plasma rifle in front of a mushroom cloud and blue lasers shooting behind it.

The week AI doom went mainstream

A new Anthropic report on Claude misuse—including people trying to use it for bioweapons research—adds to a weeklong flurry of concern around existential AI risk.

Whizy Kim

Astra’s black box problem

OpenAI’s new model reportedly doesn’t show some of its work, setting off alarms among safety researchers that AI is becoming even harder to monitor.

Whizy Kim