Are Big Companies Keeping Quiet About the Risks of AI?

Some of the world's biggest companies are embracing artificial intelligence without publicly acknowledging the risks it could bring, a new international study has found. Researchers estimate that 62% of the world's 1,000 largest companies did not acknowledge any AI risks in the public documents examined, raising questions about how transparent businesses are being with customers, employees and investors.

The research, released on 3 November 2026, examined documents from 200 major companies and incorporated assessments from 272 AI risk experts. Led by researchers including Alexander K. Saeri, Jess Graham and Michael Noetel, it investigated whether businesses publicly recognised the potential problems associated with increasingly powerful AI systems.

What did the researchers uncover?

Among the world's 100 largest companies, 32% did not publicly acknowledge any AI risks in the documents examined. The researchers estimated that the proportion increased to 62% across the world's largest 1,000 companies, although the statistical confidence interval ranged from 54% to 70%.

This is an important distinction because the 62% figure is an estimate rather than a direct count of 1,000 individual businesses. It also measures public disclosure, not whether companies have internal safety policies or effective risk-management systems.

The researchers assessed 24 categories of AI risk using the MIT AI Risk Repository, which brings together scientific evidence about potential harms associated with the technology. Five categories accounted for more than half of recorded acknowledgements, including discrimination, cyberattacks, fraud, privacy and the devaluation of human creativity.

Which AI risks are companies overlooking?

The study identified considerable differences in the concerns companies publicly recognised. Among general-purpose AI developers, 85% acknowledged discrimination risks, while only 15% acknowledged AI misalignment, a term describing situations in which AI behaviour fails to reflect its intended objectives or constraints.

Other concerns receiving limited attention included the concentration of power among AI developers and competitive pressure to introduce increasingly capable systems without sufficient consideration of safety. Researchers also found that companies operating in sectors considered particularly vulnerable to specific risks frequently failed to acknowledge those concerns publicly.

The findings are especially relevant to the creative industries, where AI is being introduced into film production, music, visual effects and digital content creation. Questions about copyright, performers' likenesses, consent and the use of creative work in training AI systems have become increasingly important to actors, musicians and writers.

Does this mean companies are deliberately hiding risks?

The research does not establish that companies are intentionally concealing information. Some businesses may have internal safeguards that were not visible in the documents reviewed, while publicly recognising a risk does not necessarily mean it is being managed effectively.

What the study demonstrates is a substantial gap in publicly available information. That gap can make it difficult for consumers, employees and investors to understand how businesses are responding to the potential consequences of AI.

For entertainment companies using increasingly sophisticated technology, public confidence may depend on more than assurances that AI is being deployed responsibly. Clear explanations of how copyright, consent and other risks are being assessed could become an important part of maintaining trust.

Research and sources

Alexander K. Saeri and colleagues, Unmitigated? How Are the World's Largest Companies Responding to AI Risks?, 3 November 2026. Read the research preprint.