AI-generated deepfakes are giving employees new ways to fabricate evidence in workplace investigations, and the technology is becoming harder to detect, warns Tracey Diamond, a labor and employment partner at Troutman Pepper Locke in Philadelphia. For HR leaders, the rise of convincing fake documents, photos, and communications means old-fashioned investigative rigor is more critical than ever-and the legal standard that protects employers hasn't changed.
"It's sort of like a continuation of the old version of employees lying. They're just creating evidence to further their lies," Diamond said. She recounted a recent investigation involving falsified communications, though she could not confirm whether AI was used. "I don't know whether the person used AI to create that, but he'd used something."
The detection gap is substantial. When identity verification firm iProov tested 2,000 U.S. and U.K. consumers in 2025, only 0.1% correctly identified every real and fake image and video sample, even after being told to look for fakes. The FBI's Internet Crime Complaint Center logged approximately $893 million in AI-enabled fraud losses in 2025, the first year the agency tracked AI as its own category.
When the evidence itself is on trial
Deepfakes create a two-sided problem. Fabricated evidence can support a false complaint, while employees accused of genuine misconduct can claim authentic evidence was manipulated. Diamond said the answer lies in rigorous, old-school investigation: checking devices on both sides, examining deleted files, and performing forensic analysis. "Do the best they can to corroborate the story," she said.
For high-stakes cases, she recommends outside help. "If it's an investigation that involves potential termination or a high-level situation that could put the company at risk, we recommend that they use outside experts to make sure they're getting it right."
The legal shield hasn't changed
Even as technology makes truth harder to verify, the employer's obligation remains a prompt and thorough investigation. If an employer reasonably believes a policy or law was violated and acts on that belief, the quality of the investigation-not the accuracy of the final conclusion-is what protects them.
"So if an employer ultimately gets it wrong, but after they conducted a thorough investigation and it was reasonable for them to reach that conclusion because the technology was that good, then that should protect them," Diamond said.
Deepfakes spread beyond the investigation room
Diamond sees fabrication creeping across the entire employment lifecycle. Gartner predicts that by 2028, one in four job candidate profiles worldwide will be fake, and its survey of 3,000 job seekers found 6% admitted to interview fraud. Visual tells in deepfake images-odd shadows, missing body parts-are fading as the technology improves.
Her larger concern is the erosion of trust. "It's hard to be able to in good faith rely on what you're seeing with your own two eyes if there's a chance that it's being faked," she said.
Policy, training, and healthy skepticism
Diamond does not believe deepfakes require a standalone policy, but they should be explicitly addressed within an employer's broader AI policy. "While you'd think it goes without saying that AI should only be used to tell the truth, maybe it warrants saying that in a written policy," she said. AI training that covers deepfake concerns should now be standard, she added, and HR leaders should work closely with their technology teams to understand both the risks and the capabilities of the tools.
Diamond also flagged the risk of employees weaponizing deepfakes against coworkers-using someone's likeness to damage a reputation or fabricate misconduct. "It's important to take a step back and think, what are the ways in which employees that want to do harm can do harm using the technology?" she said.
Why this matters for HR leaders
Diamond's advice comes down to a few practical steps: know the technology is out there, approach evidence with skepticism, and never accept what you're given at face value. "Do a little bit more digging and probing to see whether you can find corroboration for it," she said. The legal standard that protects employers-a prompt, thorough, and reasonable investigation-remains the strongest shield. Invest in forensic rigor, update your AI policies, and make deepfake awareness part of every manager's training.
Your membership also unlocks:
