As Hong Kong is learning, a deepfake crime epidemic is upon us

It all started with a routine-looking link from a securities firm. Customers who clicked on it found their trading accounts hijacked by scammers, who ended up stealing HK$46 million (US$5.86 million) in a market manipulation and phishing operation. Last week, Hong Kong police arrested eight in connection with the crime. The fraudsters had impersonated licensed brokers, stole login details, and pumped and dumped obscure stocks through stolen accounts. While phishing is not new, the scam underscored the alarming reality of identity exploitation in the digital age. As society races to contain old threats, criminals are evolving – today’s financial crime often hinges on sophisticated biometric theft, making detection even more challenging. Welcome to the age of deepfakes. In a world increasingly driven by artificial intelligence (AI), our faces, voices and even patterns of digital movements are no longer just personal identifiers but raw material. Scraped from the nooks and crannies of the internet, this data is repurposed as training material or synthetic content through a slew of affordable AI tools. We have become reservoirs of exploitable data – unprotected, commodified and with little legal recourse to reclaim what was once uniquely ours. Deepfakes show how easily fragments of our identities can be stitched together to create convincing forgeries. In the early days especially, deepfakes seemed innocuous, even entertaining. Even today, regulatory frameworks often underestimate the broader impact of such technology. The European Union’s AI Act, for example, classifies some deepfake applications as “limited risk”. This approach can obscure its real-world dangers: in the wrong hands and context, even so-called low-risk tools can inflict profound harm. This isn’t limited to celebrities or public figures. Anyone with an image can become fodder for deepfake fraud, used in ways they never consented to and may never even discover. And deepfake deception is evolving with technology. Just a year ago, deepfakes still had glitches a sharp eye could detect. Now, with technical breakthroughs like generative adversarial networks, they can produce lifelike videos and mimic voices in real time. As the technology becomes more widespread, the pool of potential victims has expanded at an unprecedented pace. This has ushered in a new wave of crime: large-scale financial fraud, which is easier to execute than before yet harder to detect. Banks and financial institutions are increasingly in the crosshairs. For example, “know your customer” checks, once the golden standard for fraud prevention, are increasingly susceptible to deepfake-altered selfies and stolen identities. Other security measures – lengthy forms and manual document checks – are little more than paper shields. Gone are the days when scammers depended on crude, cut-and-paste forgeries. Now, with just a stolen image or scraped photo, they can impersonate anyone with astounding realism. With unsettling regularity, that “anyone” could be you or me. Consider the stakes: if old methods like phishing could bamboozle people out of HK$46 million, imagine the industrial-scale chaos that can be caused by an army of AI-generated lookalikes. This epidemic is reflected in the statistics. In the first quarter of the year alone, Hong Kong recorded an astonishing 1,900 per cent year-on-year increase in deepfake-related fraud cases. This may just be the beginning. Despite the mounting threat, only 42 per cent of companies in Hong Kong are investing in AI. Of those, just 26 per cent say improving security is a top concern for the next 12 months. But the true cost of identity exploitation goes far beyond dollars and cents. It erodes trust – between individuals, within companies, across institutions and throughout society. When we can no longer distinguish real from fake, the very fabric that binds our economy and institutions begins to unravel. Some companies are taking bold steps. Meta recently sued a Hong Kong-based company whose app is accused of generating non-consensual explicit images of women. It made headlines: for once, a tech giant was taking a stand against alleged digital impersonation and image-based abuse. But Meta’s legal action is the exception, not the rule. And while tech giants can flex their legal muscles, what about the actual victims? Do they have the resources to fight back? Would they even know their faces have been weaponised and spread by strangers? The lawsuit offers a rare moment of corporate accountability, but also exposes a shameful gap: most victims don’t have an army of lawyers or the clout to demand justice. Most will never know or see justice. Thankfully, we’re still at the outset of this epidemic. Solutions should start with the basics: robust verification protocols like multi-factor authentication and out-of-band checks. Technology can also be part of the remedy. Deepfake detection software can be used to scan videos and audio for telltale signs of tampering. Content provenance tools can embed cryptographic data or watermarks at the point of creation to help confirm authenticity. Appearances, more than ever, deserve a second look. In this past week, FinVolution’s 2025 global data science competition showcased new deepfake detection technology, with the winning team achieving an impressive success rate in fake voice recognition. The true beneficiaries of these breakthroughs may be all of us. Philosopher René Descartes urged: “If you would be a real seeker after truth, it is necessary that at least once in your life you doubt, as far as possible, all things.” In this new era, questioning all we see and hear may be our best defence. It may not be ideal, but for now, it is the best we can do. For a global city like Hong Kong, the deepfake threat is not just financial. It is existential.

As Hong Kong is learning, a deepfake crime epidemic is upon us

.css-1c6uqr6{color:inherit;font-weight:inherit;font-size:inherit;font-family:inherit;line-height:inherit;overflow-wrap:break-word;}Anyone with an image can become fodder for fraud. For a global city like Hong Kong, the threat is not just financial, it is existential

Anyone with an image can become fodder for fraud. For a global city like Hong Kong, the threat is not just financial, it is existential