According to a help page posted to OpenAI‘s website last week, companies may soon need to go through an ID verification process before they can use some of the company’s future AI models.
According to the page, the Verified Organization verification procedure is “a new way for developers to unlock access to the most advanced models and capabilities on the OpenAI platform.” A valid government-issued ID from one of the nations where OpenAI’s API is compatible is needed for verification. According to OpenAI, an ID can only validate one organization per ninety days, and not all organizations will be qualified for verification.
As stated on the page, “At OpenAI, we take our responsibility to ensure that AI is both broadly accessible and used safely very seriously.” Regretfully, a tiny percentage of developers purposefully break our usage guidelines by utilizing the OpenAI APIs. As we continue to make cutting-edge models accessible to the larger developer community, we are incorporating the verification process to prevent the misuse of AI.
As OpenAI’s products grow more advanced and powerful, the new certification procedure may be meant to strengthen security around them. The business has released a number of reports detailing its attempts to identify and stop malicious usage of its models, including by purportedly North Korean-based organizations.
It might also be intended to stop intellectual property theft. A Bloomberg story earlier this year said that OpenAI was looking into whether a group associated with the China-based AI lab DeepSeek had violated its restrictions by using its API to exfiltrate vast volumes of data in late 2024, presumably for model training.
Last August, OpenAI restricted access to its services in China.