By Chuck Gallagher — Business Ethics Keynote Speaker and Trainer
TL;DR: Chuck Gallagher, an AI ethics speaker and author, explains how deepfakes have moved far beyond election ads into corporate fraud, financial scams, and image-based abuse — and why the deeper damage is the erosion of trust itself. The technology is now cheap, fast, and convincing, and the consequences land on businesses, families, and individuals who never saw it coming.
Deepfakes Have Outgrown the Ballot Box
In January 2024, a finance employee at the global engineering firm Arup joined a video call with people who looked and sounded exactly like his chief financial officer and several colleagues. Every person on that call was fake. By the end of the day he had approved fifteen transfers totaling about $25.6 million to accounts in Hong Kong, following instructions from executives who were never there. None of Arup’s systems were breached. No password was stolen. The attackers built convincing copies of real people from footage that was already public, then asked an employee to do his job.
What makes deepfakes a problem beyond politics?
Most of the public conversation about deepfakes still centers on elections, and for a while that focus made sense. But the bigger story in 2025 and 2026 is unfolding in accounting departments, family phone calls, and private group chats. Synthetic media has become cheap enough, fast enough, and realistic enough that the political angle is now the smallest part of the problem.
The numbers carry the weight. Cybersecurity firm Surfshark reported that deepfake-related scams stole roughly $1.1 billion worldwide in 2025, about three times the 2024 total. Deepfake fraud attempts have climbed more than 2,000 percent over three years. Deloitte projects that generative-AI-enabled fraud losses in the United States could reach $40 billion by 2027. A scammer now needs as little as three seconds of audio to clone a voice with roughly 85 percent accuracy, and that audio is easy to pull from a podcast, a webinar, or a YouTube clip.
How are deepfakes hitting businesses?
The Arup case is not an outlier. In July 2024, Ferrari narrowly avoided a similar attack when someone impersonated CEO Benedetto Vigna’s voice on a WhatsApp call; a suspicious executive asked a personal question the imposter could not answer, and the scheme collapsed. Chuck Gallagher, an AI ethics speaker and author, points to cases like these as proof that the threat is rarely a technical breach. It is a trust exploit. The attacker does not break the system. The attacker breaks the human assumption that the face on the screen and the voice on the line belong to who they claim to be.
That distinction matters for leaders. Arup’s global information officer described the incident as technology-enhanced social engineering rather than a cyberattack, because none of the firm’s systems were compromised. The defense, then, is not only software. It is process: verification steps, callback rules, and a culture where an employee feels safe pausing a multimillion-dollar transfer to ask one more question. The choice to build that pause, or to skip it, is the whole ballgame.
What about the damage to ordinary people?
The largest category of deepfakes has nothing to do with money or campaigns. By most estimates, roughly 98 percent of deepfake videos online are pornographic, and the overwhelming majority target women who never consented. The harm reached a breaking point with viral fake images of public figures and a wave of incidents in schools, where students generated explicit images of classmates from ordinary photos.
Congress acted. On May 19, 2025, the TAKE IT DOWN Act was signed into law, the first federal statute aimed squarely at non-consensual intimate images, including AI-generated ones. It makes knowingly publishing such content a federal crime, with up to three years in prison when minors are involved, and it requires covered platforms to remove flagged material within 48 hours starting in May 2026. The first conviction came in April 2026. The National Center for Missing and Exploited Children reported more than 1.5 million tips tied to generative AI and child exploitation in 2025 alone.
Why is the erosion of trust the deepest consequence?
There is a quieter consequence that may outlast every individual scam. Researchers call it the liar’s dividend. Once everyone knows realistic fakes exist, anyone caught on genuine video or audio can simply claim it was faked. The burden flips: instead of proving a recording is false, the accused benefits from doubt alone. This has already surfaced in courtrooms, where defendants have argued that real evidence might be AI-generated. The World Economic Forum has ranked misinformation and disinformation as its top short-term global risk two years running.
As an AI ethics speaker and author, Chuck Gallagher frames this as the heart of the matter: when people can no longer trust what they see and hear, the cost is not one fraud or one fake image. It is the shared assumption that evidence means something. And every deepfake harm traces back to a choice — to clone a voice, to post an image, to skip a verification step, to forward something without checking. The technology lowers the cost of bad choices and raises the stakes of good ones. Organizations that treat synthetic media as purely an IT issue will keep losing money and trust. The ones that treat it as a leadership and ethics issue will build the habits that hold up when the face on the screen is a lie.
Frequently Asked Questions
What is a deepfake?
A deepfake is synthetic audio, video, or imagery generated or altered by artificial intelligence to convincingly imitate a real person’s face, voice, or likeness. Modern tools can produce one from only a few seconds of source material, which is why public clips of executives and public figures are now a liability.
Are deepfakes illegal?
It depends on the use and the location. The federal TAKE IT DOWN Act, signed in 2025, criminalizes non-consensual intimate deepfakes nationwide. More than 30 states regulate AI-generated political media, and others protect likeness and voice rights. Fraud committed with a deepfake is already illegal under existing fraud and impersonation laws.
How can businesses protect themselves from deepfake fraud?
Chuck Gallagher, an AI ethics speaker and author, advises layered verification rather than gadgets alone: callback procedures on any financial request, code words for executive approvals, limits on how much executive audio and video is posted publicly, and training so employees feel empowered to pause and confirm. The $25 million Arup loss unraveled the moment one employee finally called headquarters.
Can you spot a deepfake just by looking?
Sometimes, but it is unreliable and getting harder. Detection tools help, yet the stronger defense is process and verification rather than the human eye. Asking a live caller an unexpected personal question, or requesting an unusual movement on video, still trips up many fakes today.
Bring This Conversation to Your Team
Deepfakes are not a future problem. They are a present-tense test of how organizations handle trust, verification, and the choices their people make under pressure. Chuck Gallagher helps audiences turn that test into a strategy, connecting AI ethics to the everyday decisions that protect a business and the people inside it. To bring this conversation to your team or event, visit ChuckGallagher.com.
Five Questions to Ask Your Leadership Team
1. Where in your organization could a convincing impersonation of a leader cause the most damage right now?
2. What verification step would have stopped the Arup transfer — and do you have one like it?
3. How much executive audio and video does your company post publicly without considering how it could be reused?
4. If a genuine recording of your organization were challenged as “probably fake,” could you prove it was real?
5. Are your people empowered to pause and question a request from someone who looks and sounds like the boss?
