The Midas Project

Watchtower tracks changes to corporate and government AI safety policies, both announced and unannounced. Click any entry for details.

Type:

Watchtower Corporate/Government
  • All
  • Corporate (15)

Entity:

Watchtower Entity FIlter

Importance:

Watchtower Level
  • All
  • Major (3)
  • Moderate (7)
  • Slight (5)
Watchtower Unannounced
  • Unannounced (7)

Anthropic

Corporate
Major

→

Released an updated version of their Responsible Scaling Policy.

Magic.dev

Corporate
Moderate

→

Released a statement on AI safety priorities, and announced an upcoming v2 of their responsible scaling policy.

OpenAI

Corporate
Moderate

→

Released preparedness scorecard for their newest model, o1.

Cognition

Corporate
Moderate

→

Released an acceptable usage policy, along with a reporting email for security vulnerabilities.

OpenAI

Corporate
Unannounced
Slight

→

Removed author from GPT-4o system card.