A new research project led by universities in the UK is using Sir Michael Caine's iconic voice to understand how easily people can be fooled by AI-generated audio deepfakes. The study, conducted by Dr. Ben Gibb-Reid at the University of York and Dr. Jennifer Williams at the University of Southampton, will present participants with samples of Caine's real voice alongside an AI replica created by ElevenLabs, a leading UK-based voice AI company.

Why Can't Our Ears Detect Fake Voices Like Our Eyes Detect Fake Images?

While people have become increasingly skilled at spotting AI-generated images, synthetic audio remains far more convincing to the human ear. The research team aims to identify specific linguistic features that might reveal when a voice is artificially generated, helping both the public and content creators protect themselves from malicious deepfakes.

"We are becoming increasingly familiar with the signs of AI-generated images, but the human ear is not yet as well trained at identifying synthetic audio," said Dr. Ben Gibb-Reid. "We will explore a range of linguistic features that may act as 'tells' of authenticity, and contribute to the growing need for effective voice-rights protection."

Dr. Ben Gibb-Reid, University of York

The project, funded by the YorVoice initiative at the University of York, addresses a growing concern as AI voice technology becomes more sophisticated. Caine, who has previously advocated for ethical AI use, was among the first major celebrities to support licensed AI voice technology when intellectual property and publicity rights are properly protected.

How Can This Research Protect Performers and the Public?

• Voice Rights Protection: The findings will help establish better safeguards for actors, musicians, and other performers whose voices could be replicated without consent or compensation.
• Deepfake Detection Training: By identifying linguistic markers that distinguish real from synthetic speech, researchers can help train both AI systems and human listeners to spot fraudulent audio.
• Platform Accountability: Results will inform how voice AI companies like ElevenLabs implement safeguards and work with creatives to maintain control over their identities and intellectual property.

Caine explained his motivation for participating in the study, stating that if his voice could help researchers better understand how people perceive AI-generated speech, protect artists' rights, and help defend the public from harmful deepfakes, he was pleased to support the work.

Mati Staniszewski, co-founder and chief executive of ElevenLabs, emphasized the importance of building public trust in voice AI providers as the technology advances. "At ElevenLabs, we've embedded safeguards throughout our platform, and we work closely with leading creatives who want to protect and expand their IP," Staniszewski stated. "Research of this kind will help us defend against bad actors, and keep artists in control of their voice and identity".

"If my voice can help researchers better understand how people perceive AI-generated speech, protect artists' and actors' rights, and help protect the public from harmful deepfakes, then I'm pleased to support this important research," said Sir Michael Caine.

Sir Michael Caine, Actor

The timing of this research is particularly significant. As AI voice technology becomes more accessible and realistic, the potential for misuse grows alongside legitimate applications. The study represents a collaborative effort between academic researchers, technology companies, and high-profile public figures to stay ahead of the deepfake curve. By understanding how people currently perceive synthetic speech, the research team hopes to develop better detection methods and inform policy discussions around voice rights and AI regulation.

The project underscores a broader shift in how the entertainment and technology industries are approaching AI ethics. Rather than simply deploying voice AI technology, companies and creators are increasingly investing in research to understand its implications and build protective frameworks. Caine's involvement signals that even established celebrities recognize the value of proactive engagement with AI research, particularly when it serves to protect both individual rights and public safety.