Access to future AI models in OpenAI’s API may require a verified ID

OpenAI might quickly require organizations to finish an ID verification course of with a view to entry sure future AI fashions, based on a assist web page revealed to the corporate’s web site final week.

The verification course of, referred to as Verified Group, is “a brand new manner for builders to unlock entry to probably the most superior fashions and capabilities on the OpenAI platform,” reads the web page. Verification requires a government-issued ID from one of many international locations supported by OpenAI’s API. An ID can solely confirm one group each 90 days, and never all organizations can be eligible for verification, says OpenAI.

“At OpenAI, we take our accountability severely to make sure that AI is each broadly accessible and used safely,” reads the web page. “Sadly, a small minority of builders deliberately use the OpenAI APIs in violation of our utilization insurance policies. We’re including the verification course of to mitigate unsafe use of AI whereas persevering with to make superior fashions obtainable to the broader developer group.”

OpenAI launched a brand new Verified Group standing as a brand new manner for builders to unlock entry to probably the most superior fashions and capabilities on the platform, and to be prepared for the “subsequent thrilling mannequin launch”

– Verification takes a couple of minutes and requires a legitimate… pic.twitter.com/zWZs1Oj8vE

— Tibor Blaho (@btibor91) April 12, 2025

The brand new verification course of may very well be supposed to beef up safety round OpenAI’s merchandise as they turn into extra refined and succesful. The corporate has revealed a number of studies on its efforts to detect and mitigate malicious use of its fashions, together with by teams allegedly primarily based in North Korea.

It could even be geared toward stopping IP theft. In response to a report from Bloomberg earlier this 12 months, OpenAI was investigating whether or not a bunch linked with DeepSeek, the China-based AI lab, exfiltrated giant quantities of knowledge via its API in late 2024, probably for coaching fashions — a violation of OpenAI’s phrases.

OpenAI blocked entry to its providers in China final summer season.