From bugs to bypasses: adapting vulnerability disclosure for AI safeguards

Exploring how far cyber security approaches can help mitigate risks in generative AI systems

Useful Information