🛡️Microsoft, Google, and xAI Will Let the U.S. Government Test Frontier Models Pre-Launch
Microsoft, Google, and xAI Will Let the U.S. Government Tes…
TL;DR
A CNN Business report says Microsoft, Google, and xAI have agreed to give the U.S.
A CNN Business report says Microsoft, Google, and xAI have agreed to give the U.S. Commerce Department's Center for AI Standards and Innovation early access to frontier AI models for national security testing before public release. The pact is the most concrete pre-deployment safety arrangement to date.
Key Points
Commerce Department's CAISI gets pre-launch model access.
Covers Microsoft, Google, and xAI frontier systems.
Builds on earlier voluntary commitments from the AI safety summits.
Why It Matters
Pre-deployment government testing is becoming a de facto U.S. AI safety standard, and Anthropic's notable absence from the Pentagon deal earlier this month makes the politics complicated.
Frequently Asked Questions
Why does this matter?
Pre-deployment government testing is becoming a de facto U.S. AI safety standard, and Anthropic's notable absence from the Pentagon deal earlier this month makes the politics complicated.
What happened?
A CNN Business report says Microsoft, Google, and xAI have agreed to give the U.S.
Comments
Be the first to comment
Enjoyed this article?
Get it daily. 7am. Free. Reads in 5 minutes.
Join 3,461 builders reading daily.