Anthropic Model Safety Bug Bounty Initiative (original) (raw)

Thank you for your interest in Anthropic's Model Safety Bug Bounty Initiative. This program will test a new system we've developed for AI safety mitigations. Participants will be given access to this next-generation system and challenged to identify potential vulnerabilities or ways to circumvent our safety measures. Specific criteria, instructions, and feedback will be shared with the participants of the program once they join.

Our team will carefully review all applications and will be in touch if we'd like to move forward with your participation in the program. Due to the high volume of applications we expect to receive, we regret that we may not be able to respond individually to all submissions. We appreciate your understanding.

Regardless of the outcome, we thank you for your commitment to advancing AI safety.