OpenAI's new move amid claims that DeepSeek copied its code

OpenAI now requires government ID verification for developers who want access to its most advanced AI models in a possible bid to keep its models out of the reach of Chinese competitors.

This comes amid a new research paper from AI content detection company Copyleaks estimating that 74.2% of the outputs from rival Chinese AI model, DeepSeek-R1, were classified as written by OpenAI.

Also Read: OpenAI's Sam Altman announces hiring: ‘If you have a background in…’

“This discovery raises concerns about DeepSeek-R1’s resemblance to OpenAI’s model, particularly regarding data sourcing, IP rights, and transparency,” Copyleaks wrote in an official statement. “Undisclosed reliance on existing models can reinforce biases, limit diversity, and pose legal or ethical risks.”

OpenAI said in its website that ID verification has been brought because “a small minority of developers intentionally use the OpenAI APIs in violation of our usage policies.”

However, once verified, developers will have access to Reasoning Summaries from o1, o3-mini, o3 and o4-mini.

OpenAI also said that it supports “identification from over 200 countries.” However, it hasn't given an open list and simply states that “the best way to see if your country is supported is to start the verification process.”

Copyleaks says that it combined three AI classifiers, all trained on texts from Claude, Gemini, Llama, and OpenAI which “identified subtle stylistic features like sentence structure, vocabulary, and phrasing” which it claims has a 99.88% precision rate and just a 0.04% false-positive rate.

Also Read: Airtel SIM cards delivered in 10 mins on Blinkit now. Check details here

The classifier, which was also tested on other models such as Microsoft's phi-4 and Elon Musk's Grok-1 found that those models had almost zero similarity to OpenAI, scoring 99.3% and 100.3% "no-agreement" respectively, which means they were trained independently.

Mistral's Mixtral model had some amount of similarities, but DeepSeek's numbers stood out starkly, according to the report.

All of this comes after DeepSeek spread shockwaves across the AI community by releasing its model earlier this year, which performed similar to Silicon Valley offerings, but at a fraction of the cost.

Copyleaks meanwhile, had also mentioned in its statement that “DeepSeek’s claims of a groundbreaking, low-cost training method—if based on unauthorized distillation of OpenAI—may have misled the market, contributing to Nvidia's $593 billion single-day loss and giving DeepSeek an unfair advantage.”

Also Read: Google search engine's URL is changing. Here's what it will look like

There is also another side to this picture. OpenAI itself built its early models by scraping the web, using content from news publishers, authors, and creators, often without consent, leading to multiple lawsuits.