OpenAI Faces Scrutiny Over One-Week Safety Testing for GPT-4 Omni

Jul 12, 2024, 08:15 PM

OpenAI is facing significant scrutiny over its recent handling of safety protocols for its latest AI model, GPT-4 Omni. Despite promising the White House last year to rigorously test new versions of its technology for safety, reports indicate that OpenAI's safety team felt pressured to expedite the testing process to meet launch deadlines. Some employees have voiced concerns that the company failed to uphold its safety commitments, with safety testing being crunched into just one week. OpenAI, however, maintains that it did not cut corners on safety. The situation has led to growing criticism and has been described by some as 'defensive window dressing' in response to these concerns.