Imagine scrolling through your social feed and seeing a familiar face — a beloved actor, a trusted athlete, a favorite influencer — telling you about an amazing new product, an investment opportunity, or a giveaway that seems too good to pass up. The video looks flawless. The voice sounds authentic. The mannerisms are spot-on. And yet, none of it is real. This is the new reality of the digital age, where generative artificial intelligence has made it possible to create videos and images so convincing that even the people closest to the person on screen can’t tell the difference. These are called deepfakes, and they are no longer a far-off technological curiosity. They are a present and growing threat to trust, credibility, and the entire economy of influence and identity. Trust has always been the foundation of commerce and relationships, but in today’s digital ecosystem, trust is not just a soft value — it is operational infrastructure. When people lose faith in what they see and hear, particularly when it involves the faces and voices of celebrities, creators, athletes, and the brands that work with them, the pillars of marketing and personal brand value begin to crumble. The danger is no longer hypothetical. Generative AI is now accessible to almost anyone, and it has opened the door to a world where fake endorsements, cloned social media accounts, and entirely fabricated influencers are becoming a legitimate and increasingly common vector for fraud.
Some of the most visible examples have already made headlines and forced the public to take notice. Tom Hanks, one of the most beloved actors in Hollywood, took to Instagram to warn his followers about an AI-generated version of himself that was appearing in a video promoting a dental plan. He was not involved, he said firmly, and urged people to beware. The video looked so real that it could have easily fooled anyone who wasn’t paying close attention. Around the same time, YouTube star Jimmy Donaldson, better known as MrBeast, found himself at the center of a deepfake giveaway scam that circulated across platforms. In it, a convincingly realistic version of him appeared to be offering free iPhones to unsuspecting fans. MrBeast publicly questioned why platforms like X were not doing more to stop the spread of such content. These are the high-profile, publicly known cases — the ones that generate news coverage and outrage. But the same technology is also being used quietly and relentlessly against NIL athletes, mid-tier creators, and agencies that manage six- and seven-figure brand deals. The attack surface has expanded dramatically. What was once a tool for celebrity impersonation has become a scalable method of deception aimed at anyone with a digital following and financial value.
The tools required to create these sophisticated fakes are no longer limited to elite researchers or well-funded intelligence agencies. Publicly available generative AI platforms now allow anyone with basic technical knowledge to create voice clones trained on just a few seconds of public audio, swap faces with startling precision, manipulate lip movements, and even construct synthetic personas that have never existed in the real world. Adversaries can scrape training data from podcasts, Instagram Reels, LinkedIn posts, conference talks, and TikToks, compiling a rich profile of a person’s appearance and voice without ever coming into direct contact with them. The result is a form of industrial-scale social engineering that targets the creator and athlete-marketing economy in ways that go far beyond isolated celebrity antics. Perhaps most troubling is the combination of deepfake content with account takeover. When a hacker gains access to an official channel and then releases deepfake content from that verified account, the deception is magnified exponentially. Followers trust the account because it is official, and brand partners trust it because it has always represented the talent. The fraudulent post appears practically indistinguishable from legitimate content, and by the time the truth comes out, the damage has already spread across the internet.
To understand the depth of this threat, it helps to hear from someone on the front lines. Jessica Bartlett is the founder and CEO of Verge Agency, a Los Angeles-based talent management company that specializes in influencers and NIL athletes. She has seen firsthand how quickly digital identities can be compromised and how unprepared many agencies are to respond. In her view, the industry is only beginning to wake up to the reality of deepfakes. Her firm has developed strict account-security procedures designed to protect client brand deals, and she recommends that all agencies, at a minimum, enforce strong authentication requirements, maintain controlled access to accounts, document clear recovery procedures, continuously monitor for impersonation and cloned accounts, and establish a defined escalation process for suspicious content. What she emphasizes most is that verification in the age of AI can no longer rely on simple recognition. Seeing a person’s face in a video is no longer proof that they are speaking. Hearing a familiar voice is no longer proof that they endorsed a product. Verification has to go much deeper, involving a layering of technical checks and human judgment. Bartlett also believes the industry is moving toward a more formalized approach. She has said that what is needed is a shared security standard, not just a scattered collection of informal best practices. As she puts it, these expectations will increasingly show up in agency agreements, brand contracts, insurance requirements, and due diligence processes. When an influencer’s digital identity can be worth millions of dollars, protecting account integrity is no longer optional.
The scale of the problem is staggering and growing quickly. Deepfake production is now estimated to be nearly one million videos per month, and that number is rapidly accelerating. What was once a niche area of academic research has become a mainstream risk. There have already been documented cases of multi-million-dollar wire transfers being approved after staff members joined video conferences with deepfake versions of their own executives, convincing enough to authorize major financial transactions. Complex impersonations have targeted public and private sector figures, and the same toolkit that enables business email compromise is now being used in talent marketing. Deepfake videos appear to show real creators endorsing products or investment schemes they never touched. Cloned accounts post fake brand endorsements designed to trick fans. Fictitious influencer personas who never existed in real life build the appearance of a following and then capitalize on that fabricated trust. The consequences for talent and brands are immediate and intensifying. A fraudulent endorsement can shatter follower trust, trigger contractual disputes, cause serious reputational harm, and expose the person whose likeness was stolen to legal liability. For agencies that manage the portfolios of athletes and influencers, a single compromised account or undiscovered deepfake campaign can put substantial revenue at risk. And yet, despite the stakes, most agencies still treat social account security as basic hygiene rather than a fundamental operational control. There are no standard operating procedures for multi-factor authentication, recovery protocols, monitoring for synthetic media, or clear escalation pathways when something suspicious appears. In this new landscape, seeing is no longer believing. Verification is the new standard.
The way forward requires practical action built on the timeless principle of “trust but verify.” Identity should be confirmed out-of-band before responding to high-stakes requests or content. A voice message, a video call, or a social media post should never be enough to trigger a significant decision or transfer. Any new or unusual digital relationship, and any sudden shift in posting behavior, should be treated as potentially artificial until it is proven otherwise. Robust account security measures should be implemented across all platforms, including phishing-resistant multi-factor authentication, passkeys, unique credentials, limited administrator access, ongoing monitoring, and quick recovery playbooks. Zero Trust principles should be applied not only to networks but also to human interactions and talent representation. Teams should be trained to recognize deepfake indicators, such as unnatural mannerisms, odd prosody, inconsistent lighting, and delayed responses, even while acknowledging that these cues are disappearing as models become more sophisticated. There is also a growing need for platform accountability. Social media companies must do more to remove fake content quickly and make provenance signals more transparent. Investing in detection tools, behavioral analytics, and responsible AI governance is essential for keeping defensive capabilities up to date with offensive ones. Cybersecurity can no longer be treated as a back-office technical concern. For brands, talent managers, and creators, it is a business risk that belongs at the board level and in the office of agency leadership. When digital trust is treated as infrastructure and protected with the same rigor as financial controls and intellectual property, organizations will be far better equipped to navigate a world saturated with synthetic media.
Jessica Bartlett reinforces this reality with a clear and urgent message. She believes that creators need an agency that has the resources to detect these issues quickly, escalate directly with platforms, preserve evidence, coordinate with brand partners, and involve experienced legal counsel immediately when necessary. Speed is critical, she says, because the longer fraudulent content remains online, the greater the potential financial and reputational damage. At Verge Agency, protection is approached at the portfolio level. An agency should understand each talent’s official accounts, revenue streams, brand partnerships, and potential points of vulnerability long before an incident happens. That means account security, verification, impersonation monitoring, legal response, and recovery procedures must be built into the core operating infrastructure of an agency, not treated as responsibilities that only come into play after something goes wrong. This also changes what creators should expect from their representation. As the industry becomes more sophisticated, choosing an agency should increasingly involve evaluating not only who can negotiate deals or grow revenue, but also who has the infrastructure to protect the business being built. Dedicated operational resources, strong platform relationships, experienced legal counsel, clear escalation procedures, and the ability to respond immediately when a creator’s identity is compromised all matter just as much as a talent agent’s ability to close a deal. As the sophistication of fraud has increased, the sophistication of those who represent talent must increase as well. The convergence of AI with other cutting-edge technologies will continue to transform privacy, security, and trust in the acceleration era. Cloned influencer accounts and deepfake endorsements are both a warning and a test. Talent and brands that make verification a standard practice will protect their value and their reputations. Those that do not will eventually discover, often too late, that the most convincing face on the screen might not actually be real.