AI companies are employing tactics similar to those used by the tobacco, pharmaceutical, and oil industries to influence legislation. A recent study demonstrates how these organizations shape narratives and policies during key global AI events. Researchers from the University of Edinburgh, Trinity College Dublin, and Carnegie Mellon University examined a range of news articles connected to significant AI discussions occurring between 2023 and 2025.
The analysis uncovered 27 distinct methods categorized as “corporate capture,” where corporate interests overshadow public welfare within regulatory systems. By reviewing 100 articles associated with major AI events, including trilogues for the EU AI Act and global summits in the UK, South Korea, and France, the researchers identified 249 instances of regulatory manipulation.
A key tactic highlighted was “narrative capture,” where companies create persuasive narratives aimed at swaying policymakers. A common assertion in these discussions was the idea that “regulation stifles innovation,” framing regulatory measures as unnecessary and excessive, frequently leading to calls for their removal.
Another notable tactic is the “elusion of law,” which involves misleading interpretations and breaches of existing regulations, including antitrust, privacy, copyright, and labor laws. The study revealed that major AI firms actively resist regulatory oversight, utilizing strategies such as lobbying and intimidating whistleblowers, researchers, and lawmakers to maintain their market power.
The research also pointed to the concerning “revolving door” phenomenon, where former government officials obtain lucrative positions in leading AI companies. This situation raises ethical questions regarding the integrity of regulatory processes. Furthermore, significant financial contributions from large AI firms to political parties, along with public officials holding stakes in these companies, complicate the regulatory environment.
The insights from this study echo lessons learned from other powerful industries, such as tobacco and pharmaceuticals. The researchers advocate for establishing clear separations between public interests and private enterprises to reduce the risks of corporate capture. Implementing binding regulations on interactions between government entities and the AI sector could help manage potential conflicts of interest.
Dr. Zeerak Talat, a Chancellor’s Fellow at the University of Edinburgh’s School of Informatics, commented on the implications of the findings: “It’s remarkable how the findings relate to common experiences of companies having greater influence over democratic processes than people.” While the study does not establish a direct causal link between corporate capture and citizen disenfranchisement, it suggests a concerning correlation.
Dr. Abeba Birhane from Trinity College Dublin’s AI Accountability Lab emphasized the recurring narratives employed by Big AI. She noted, “In addition to ‘narrative capture’ and the violations and contentious interpretations of antitrust, privacy, copyright and labor laws that were most recurrent, we also found that Big AI frequently uses the notion that ‘regulation stifles innovation’ to rationalize their control of the overall narrative.”
The findings will be presented at the upcoming ACM Conference on Fairness, Accountability, and Transparency in June 2026, marking a significant moment for discussions regarding AI regulations. A preprint version of the paper is already available on the arXiv platform, encouraging further examination and dialogue within the tech community and beyond.
The implications of this research extend far beyond academia. As AI increasingly impacts society, understanding the relationship between industry and regulation becomes essential. The pressing question remains: how can we ensure that regulations prioritize the public good over corporate interests?
