The business increased limits on research into AI concerns, according to former OpenAI staffers who departed

OpenAI is reportedly becoming more cautious about releasing research that highlights the potential risks and economic downsides of Artificial Intelligence. Several senior executives have allegedly resigned due to the tightened controls the creator of ChatGPT is placing on research into AI risks. According to a recent report by Wired, former OpenAI employees say it has become increasingly difficult to publish studies examining how AI may negatively affect jobs, markets, and broader economic stability.

The company’s economic research team has lost at least two researchers in recent months, including researcher Tom Cunningham. He reportedly expressed dissatisfaction with the team’s work being shaped by internal pressures, which led to his departure in September. In an email to colleagues, allegedly reviewed by Wired, he said the group faced “a growing tension” between conducting rigorous economic analysis and functioning as an advocacy arm.

According to the report, citing sources familiar with the matter, OpenAI has become more reluctant to publish studies highlighting potential labor disruption or job displacement. While the company previously released regular research on how AI would transform work—such as a widely cited 2023 paper on automation exposure—former employees claim there is now less interest in research that casts AI in a more cautionary or disruptive light. Some researchers believe the company increasingly favors studies that emphasize productivity gains or efficiencies rather than those that raise concerns.

OpenAI leadership, however, disputes these claims. They argue that the economic research team’s mandate has expanded, not narrowed, and that its purpose remains to assess both the benefits and drawbacks of the rapid global adoption of AI. Internal communications suggest that executives believe the organization must not only identify challenging issues but also help develop solutions.

The allegations of OpenAI suppressing or reducing research on AI’s potential downsides come at a time when the company is strengthening multibillion-dollar partnerships with governments and major enterprises. As AI technologies become deeply integrated into economies and workplaces, scrutiny is increasing over how AI labs assess and reveal the societal impacts of their systems. Critics warn that if leading companies choose to spotlight only positive findings, the public may be left with an incomplete understanding of the risks AI could pose.

Some economists who have collaborated with OpenAI claim that the company’s recent research appears increasingly framed to highlight the technology’s benefits, deprioritizing more cautious or disruptive perspectives.

Those who departed the company cite intellectual freedom as their primary reason. Cunningham’s resignation reportedly stemmed from what he saw as growing difficulty in publishing high-quality, unfiltered research. Others familiar with the team privately echo these concerns, arguing that the space for frank analysis is shrinking as the company’s political and commercial stakes continue to rise.