The Deepfake Era Has Arrived and PR Is the Front Line

Evan Nierman

In March, the National Republican Senatorial Committee released an 85-second attack ad against Democrat Texas Senate candidate James Talarico. In the ad, Talarico appears sitting before a Texas flag, reading years-old social media posts back into the camera and ad-libbing approvals of his own words.

The tweets were real, but the rest was not. Talarico never filmed the video. Campaign committee staff generated it using AI.

Hany Farid, a UC Berkeley digital forensics professor, reviewed the ad and told CNN that outside of a slight audio sync issue, the video was hyper-realistic enough that most people watching wouldn’t immediately recognize it as fake. The ad didn’t hit cable news as a scandal, but was assembled as a strategy. And it’s one of a growing list of deepfake-driven political ads already in circulation during the 2026 midterm cycle.

A month before the Talarico video, AFP fact-checkers flagged a clip overlaying AI-generated audio onto genuine World Economic Forum footage to make Donald Trump sound like he was mocking former California Governor Gavin Newsom. Reuters, in its own reporting on 2026 midterm deepfakes, cited Newsom as someone who has frequently employed deepfake videos to troll Trump.

The AI tools are in circulation across the political spectrum. So, too, is the willingness to deploy them, and the Internet can serve as a force multiplier.

No federal law directly addresses the broader threat posed by deepfakes. That reality, more than any single clip, is the story PR pros need to consider right now. Bear in mind that this phenomenon is hardly confined to politics or to any one country.

On February 26, Reuters also reported that the Bank of Italy had issued a public warning about fake articles, images and videos circulating online showing Governor Fabio Panetta appearing to endorse investment products. The central bank filed a complaint with judicial authorities, citing the need to protect the public and to safeguard both the institution and Panetta’s reputation.

Think about what that means and the kind of PR crisis it caused. The head of a G7 central bank had his face and voice weaponized to impersonate endorsements of investment products on a grand scale. The Bank of Italy had to go to court.

Any public figure is now a viable target for AI-enabled reputational manipulation, from the politician in the primary, to the celebrity on tour, to the CEO on an earnings call, to the influencer who just hit one million followers.

For years, deepfakes lived on the horizon of our industry. While a real and emerging concern, they were also the kind of risk leadership nodded along to in crisis planning sessions and set aside for a day when their power would possibly become more potent. That era is here. In the first four months of 2026 alone, deepfakes have crossed the line from threat to effective tool, requiring PR acumen and action to combat them.

The state laws in place in the United States are a patchwork and most only apply in the final weeks before an election. Europe is moving faster, with active scrutiny of AI platforms under the Digital Services Act and draft legislation advancing in Spain, but enforcement is uneven and the technology is outpacing every regulatory body trying to catch up to it.

And here is the ugly truth for professional communicators: Even when a deepfake gets flagged as a forgery, the underlying smear it was engineered to deliver often travels further because of the coverage surrounding the revelation of the deepfake.

Courts, regulators and campaigns are each moving on their own clocks, but the signals they’re sending are the same: Synthetic media is now a tool that demands consequences. Our clients are living during an era of reputational shift, whether they know it or not.

So, what does it mean for the PR work itself? The implications are staggeringly broad. Any client with a public footprint is a viable target, and the effort required to hit them is minimal thanks to breakneck pace of AI innovation.

This means that speed matters and often wins. The realism the researcher Farid described in the Talarico ad is precisely why the detection gap matters so much. The window between a deepfake going live and an authoritative rebuttal catching up to expose it is the window where reputational damage actually happens.

Teams that can mobilize inside that sliver of opportunity, the ability to authenticate or debunk the content and possessing pre-existing relationships at media outlets will hold the edge. Teams that are still building their deepfake playbooks mid-crisis will not.

The attack surface of this content is also much wider than that accounted for by most crisis plans. The risk isn’t just the CEO whose voice might be cloned for a wire transfer scam. But also the mid-level spokesperson whose LinkedIn videos become training data. It’s the board member whose keynote audio gets fed through a voice model. It’s the brand ambassador whose likeness is now cheap and available to anyone with a grudge and a prompt.

It also means legal and comms teams need to be working in concert from minute one. The TAKE IT DOWN Act, signed last year, created federal criminal penalties for nonconsensual synthetic intimate imagery and requires platforms to pull reported content within 48 hours.

It doesn’t reach political or reputational deepfakes on its own, but with convictions now on the books and DSA enforcement accelerating in Europe, the tools available in an AI-driven crisis include legal instruments that did not exist even 18 months ago. A communications response that ignores them falls short. A legal response that ignores the reputational clock does the same.

The antiquated instinct to stay quiet and let a false claim burn itself out is a dangerous one in this modern-day environment. Synthetic content lands like evidence for most audiences, because video and audio have long carried the weight of proof. After all, we’ve been programmed to believe that “seeing is believing” and that “a picture is worth 1,000 words.”

Crisis firms working in this complex space, including my own, are now facing down matters that would have been unimaginable just two years ago. Cloned voices used to manufacture controversies. Fabricated footage used to seed stories with reporters. The core tenets of crisis management may still apply, but the tools and stakes around those instincts have fundamentally changed, and they’ll be tougher than ever to combat.

We’re still early in this era, but the direction isn’t ambiguous. The courts are issuing convictions. The regulators are issuing fines. The campaigns are using the technology as live ammunition.

For communications professionals, the implication is simple to state, but hard to operationalize: A client’s reputation is now defended not only against what is said about them, but also against what can be manufactured to appear as if they said it themselves.

Reputation has always been hard to build and easy to lose. But the AI deepfake era means reputation will be easier to forge and harder to reclaim. That’s the challenge ahead and the reality with which crisis communicators must grapple.

***

Evan Nierman is the Founder and CEO of Red Banyan, a global reputation management firm, and author of Amazon bestsellers “Crisis Averted” and “The Cancel Culture Curse.”