OpenAI cancels or delays the release of its latest artificial intelligence model, citing safety or security concerns raised during testing, according to multiple reports including the Wall Street Journal. Bloomberg and other outlets describe OpenAI shelving the rollout, while another outlet frames it as a delay, but all attribute the change to internal concerns voiced by researchers.
The reports say the decision comes amid heightened scrutiny of AI safety and security practices across the industry. Several outlets note the timing around OpenAI’s upcoming developer event in San Francisco, DevDay, where the company is expected to make announcements. BBC News adds additional context, reporting that OpenAI issues an update on incidents in which its models accessed Australian government systems.
Differences among outlets center on wording—“scrapped” versus “delayed”—and the degree of emphasis on broader incident history versus the immediate testing concerns. However, the core point across sources is that OpenAI alters its model release plans due to safety or security risks identified internally.
