OpenAI introduced a new framework for tracking, investigating and publicly disclosing instances of model misalignment.

OpenAI releases six reports on unexpected model behavior under a new framework for tracking, investigating, and publicly ...

Startup says it’s learned from these mistakes and that they shouldn’t happen again … which is just what Zuck has said about ...

OpenAI published a framework for tracking, investigating, and disclosing instances of model misalignment on September 16, ...

OpenAI reveals six cases where AI models hid mistakes, invented data, and bypassed security controls as part of new ...

Imagine asking an AI for earnings figures and getting an answer based on data it was never authorised to access. OpenAI says ...

Monitoring and incident management in corporate AI operationsHello everyone. I am Daichi Sasagawa, a freelance engineer and IT consultant.These days, problematic AI behavior is a hot topic.On ...

An unreleased OpenAI model was found giving itself secret instructions where the model claimed that it was equal to humans ...

OpenAI has disclosed six incidents involving unexpected or concerning behavior by AI models as it introduces a new framework ...

OpenAI discloses six cases of AI misalignment, including models that hid errors, invented data, and bypassed restrictions, as ...

OpenAI has published six reports detailing model behaviour that raised safety and alignment concerns during training and ...

OpenAI has introduced a new framework for tracking, investigating, and disclosing model misalignment. The company has also ...