OpenAI Researcher Quits Over Advocacy

OpenAI chief global affairs officer Chris Lehane.

Introduction: A Departure That Speaks Volumes

In a development that has sent ripples through the artificial intelligence community, a researcher from OpenAI’s economic research team has resigned, alleging that the company’s approach to studying AI’s economic impact is shifting from objective analysis to advocacy. This departure, coupled with claims from four sources close to the situation, suggests that OpenAI may be becoming increasingly reluctant to publish research that highlights the negative consequences of artificial intelligence on the economy. The company, however, offers a different narrative, stating that it has merely expanded the scope of its economic research team rather than altered its focus.

The Allegations: A Shift in Research Direction

Tom Cunningham’s Departure

The central figure in this controversy is Tom Cunningham, a researcher who was part of OpenAI’s economic research team. According to sources familiar with the situation, Cunningham left OpenAI entirely in September after concluding that it had become difficult to publish high-quality research. In a parting message shared internally, Cunningham reportedly wrote that the team faced a growing tension between conducting rigorous analysis and functioning as a “de facto advocacy arm” for OpenAI ¹.

Cunningham’s concerns were not isolated. At least two researchers have reportedly left the economic research team due to what they perceived as increasing restrictions on publishing studies that might cast AI in a negative light, particularly regarding its economic implications ².

The Broader Claims

The four sources close to the situation claim that OpenAI has become more guarded about publishing research that highlights the potentially negative impact that AI could have on the economy. This shift, they argue, has contributed to recent departures from the economic research team. The sources suggest that there’s an increasing pressure within the company to present AI’s economic impact in a more favorable light, potentially compromising the objectivity that should be at the heart of academic research ¹.

OpenAI’s Response: Expansion vs. Restriction

Official Statements

In response to these allegations, OpenAI has maintained that the economic research team’s focus hasn’t shifted toward advocacy. Instead, the company claims it has only expanded the team’s scope. This distinction is crucial but perhaps semantic – expanding the scope could still result in a change in emphasis or direction, especially when commercial interests are involved ¹.

OpenAI’s economic research efforts are currently managed by Ronnie Chatterji, who serves as the company’s first chief economist. Chatterji, who joined OpenAI after serving in various capacities in the Biden administration, led a significant September report on how people around the world are using ChatGPT. Interestingly, Cunningham was listed as an author on this report, suggesting that despite his eventual departure, he was involved in significant research projects at the company ¹.

Recent Appointments and Commercial Focus

The timing of these allegations coincides with OpenAI’s increasing commercial ambitions. The company recently appointed Denise Dresser, the former CEO of Slack, as its Chief Revenue Officer. This move signals a strengthening of OpenAI’s commercial strategy as it seeks to monetize its AI technologies more effectively ³.

With this increased focus on revenue generation, questions naturally arise about how commercial pressures might influence research priorities. The departure of researchers who claim the company is becoming more restrictive about publishing negative findings suggests a potential conflict between the pursuit of profit and the commitment to objective research ¹.

Broader Context: AI Research in the Commercial Sphere

Industry-Wide Concerns

The issues raised by Cunningham and his colleagues are not unique to OpenAI. Concerns about publication restrictions and advocacy in AI research have emerged at other major AI companies as well. For instance, Google’s DeepMind has reportedly held back research that showed its Gemini language model was less capable than rivals ⁴.

This pattern suggests a broader challenge in the AI industry: how to maintain research integrity while operating as commercial entities with significant financial interests. The tension between academic freedom and corporate objectives appears to be a systemic issue rather than one confined to a single company ⁴.

Academic Freedom in Corporate Research

Academic freedom, traditionally understood as the principle that researchers should be able to pursue knowledge and share findings without undue restriction, becomes complicated when applied to corporate research settings ⁵. In universities, researchers are protected by tenure and institutional commitments to free inquiry. In commercial AI companies, however, research priorities may be influenced by market considerations, competitive positioning, and the need to maintain positive public relations.

The potential erosion of research independence in commercial AI organizations raises significant concerns about public trust in AI research. If major AI companies are perceived as prioritizing advocacy over objectivity, it could undermine confidence in their findings and recommendations, particularly when those findings inform public policy or business decisions ⁵.

Implications for AI Governance and Public Trust

The Challenge of Objective Research

The allegations against OpenAI touch on fundamental questions about how AI research should be conducted and communicated. As AI becomes increasingly integrated into economic systems, employment structures, and daily life, objective research on its impacts becomes more crucial. If companies are filtering or restricting research that shows negative outcomes, the public and policymakers may not have access to the complete picture needed for informed decision-making ¹.

This concern is particularly acute given the significant influence that major AI companies now wield in public discourse and policy development. When research findings are perceived as biased or incomplete, it can lead to misguided policies, inadequate preparation for AI’s societal impacts, and a general erosion of trust in technological institutions ².

Comparative Industry Practices

The situation at OpenAI also highlights the need for clearer standards and transparency in corporate AI research. How do different companies approach the publication of research that might show their technologies in a negative light? What safeguards exist to protect research independence? And how can the public distinguish between objective research and corporate advocacy?

Some companies, like Anthropic, have positioned themselves as public benefit corporations with explicit commitments to safety and transparency. However, even these organizations face questions about their research practices and publication decisions. The challenge appears to be structural rather than limited to any single company’s culture or policies ⁶.

Conclusion: Navigating the Future of AI Research

The departure of Tom Cunningham from OpenAI’s economic research team, along with allegations of increasing restrictions on publishing negative findings, raises important questions about the integrity of AI research in commercial settings. While OpenAI maintains that it has only expanded its research scope rather than shifted toward advocacy, the concerns raised by former employees suggest a potential conflict between commercial interests and research objectivity.

This issue is not merely about one company’s practices but reflects broader challenges facing the AI industry. As AI technologies become more powerful and pervasive, the need for objective, transparent research becomes more critical. The public, policymakers, and businesses need access to complete and unbiased information about AI’s impacts to make informed decisions.

The situation also underscores the importance of developing clearer standards for corporate AI research and mechanisms for ensuring transparency and independence. Whether through industry self-regulation, third-party oversight, or new institutional arrangements, the AI community must find ways to maintain research integrity while operating in commercial environments.

Ultimately, the credibility of AI research and the trustworthiness of AI companies depend on their commitment to transparency and objectivity. As the industry continues to evolve, how companies balance commercial interests with research integrity will likely become an increasingly important factor in their reputation and public acceptance.