FOR IMMEDIATE RELEASE

October 7, 2026

Press Contact:

Vincent Errichetti

(617) 922-1824

Press@johndeatonforsenate.com

Senate nominee calls for mandatory AI disclosure, protections for seniors and children, and real accountability for companies that deceive

BOSTON, MA. -- John Deaton, nominee for U.S. Senate in Massachusetts, today called on Congress to pass a national AI standard with enforceable consumer protections. The call follows this week's announcement of an AI video model that its maker says nearly half of test subjects mistook for a real person.

AI company Tavus unveiled the model, called Griffin, on October 1. The company reports that 48% of participants in its study believed they were talking to a real human after a live one-minute video call. It says earlier systems fooled fewer than 3%. Tavus itself acknowledges the model can deceive people into believing it is not AI.

"When a machine can look you in the eye on a video call and half the people talking to it think it's human, the scam writes itself," Deaton said. "A grandmother gets a call from a face and a voice she trusts. Minutes later her life savings are gone. That is where this goes if Washington does nothing."

"I've spent my career taking on corporations that knew their products would hurt people and sold them anyway," Deaton continued. "I'm not waiting for the body count this time. Fraud rings will use this against seniors. Lonely and vulnerable people will be sold relationships with people who don’t exist. Isolated or lonely children battling depression or other mental health issues will be especially vulnerable.”

Deaton's national standard would:

• Require disclosure. Any AI in a live video or voice conversation must identify itself as AI, up front and whenever asked.
• Ban impersonation. No AI may use a real person's face or voice without that person's consent.
• Protect children. No romantic or companion AI products for minors, backed by age verification.
• Protect seniors. Banks and payment apps must flag and verify large, unusual transfers before the money leaves.
• Hold companies accountable. Firms that deploy undisclosed human-passing AI answer for the harm, with FTC and state attorney general enforcement and a right for victims to sue.
• Set one clear rule. A single national standard, so protection doesn't depend on which state you live in.

"I'm for innovation. I've fought regulators who tried to strangle it," Deaton said. "But there is a bright line between building tools that serve people and building machines designed to pass as people. If it's AI, it has to say so. That's not anti-technology. That's basic honesty, and it should be the law."

###