How AI is Fuelling a New Wave of Online Fraud Ahead of Valentine's Day

With Valentine's Day just around the corner, romance scams are once again in the spotlight, but this year there is a dangerous new twist. Artificial intelligence is enabling fraudsters to create fake profiles, generate deepfake videos, and maintain convincing conversations at a scale never seen before. According to Crime World, 88 people in Ireland were conned out of a staggering €2.8 million by online romance fraudsters in 2025 alone.

The Scale of the Problem in Ireland

According to exclusive figures from An Garda Síochána published by Crime World, more than €9 million has been stolen from victims of romance fraud in Ireland over the past five years. Of the victims who have come forward to Gardaí since 2021, the majority were female and their average financial loss was €27,000. The highest reported loss by a woman in Ireland was over €450,000, stolen across 18 transactions, while one male victim had more than €380,000 taken.

Those figures almost certainly understate the true scale. As Crime World reported, most victims never come forward because they feel too ashamed to admit what has happened, even to themselves.

How AI Has Changed the Game

Gone are the days when a poorly written message and a stolen photo were the hallmarks of a romance scam. According to a report by credit agency Experian, AI-powered romance scams are among the top five fraud trends to watch in 2026, with bots now capable of responding convincingly, building trust over time, and manipulating victims with precision.

The 2026 Norton Insights Report found that nearly half of current online daters in the US had been targeted by a dating scam, with over 17 million dating scam attacks blocked in the final quarter of 2025 alone, representing a 19% increase on the previous year.

Research by McAfee found that more than one in four people surveyed said they, or someone they know, had been approached by an AI chatbot posing as a real person on a dating app or social media. Scammers are now using AI to generate entirely fictitious profile photos that look completely real, produce deepfake video messages, clone voices from short audio clips, and maintain multiple fake romantic relationships simultaneously through chatbots.

According to security researchers at CanIPhish, fraudsters now have access to AI-powered face-swapping technology, which allows them to assume any identity on live video calls, meaning even a video chat is no longer a reliable way to confirm someone is who they claim to be.

Speaking to Crime World, US-based dating coach Susan Winter warned that detection is becoming increasingly difficult. In earlier years, scammers would avoid video calls with excuses about being stationed on an oil rig or having poor internet. Now, she said, they can take an image of any person and make it speak convincingly, or even hire actors for virtual conversations.

The Platforms Being Targeted

While dating apps remain a key hunting ground, scammers are also posing as credible professionals on platforms like LinkedIn, as well as targeting people through Facebook, TikTok, and Instagram, according to Crime World. Fake AI-generated Instagram accounts, featuring attractive profiles with no real history behind them, have become a particularly common entry point.

According to security firm OmniWatch, romance scams rose 20% in the first quarter of 2025 compared to 2024, driven in part by the ease with which AI tools can create convincing fake identities at scale.

How the Scams Work

The pattern is well-established but devastatingly effective. Fraudsters typically use stolen photos, or now AI-generated images, to set up a fake profile. They identify potential victims and groom them over a sustained period.

According to Garda advice, their profile usually depicts a person with a responsible job but who is fictitiously based somewhere that makes them unable to travel freely or maintain reliable phone and internet access, such as military deployment overseas or remote work on an energy platform.

As reported by Crime World, the requests for money start small at first, and the scammer may even repay these sums to build trust. Over time, the requests increase, and victims find themselves pressured to hand over funds for anything from medical bills and customs duties to flights and visas.

In one case highlighted by Crime World, two people received suspended sentences in November 2025 for acting as "money mules" in a scam that conned a 90-year-old Irishman out of almost €25,000. The victim had connected with a woman he believed to be a US navy chef on Match.com, and was convinced to part with money to cover inheritance storage fees linked to a fabricated €31 million estate in Dubai.

What Gardaí Have Said

In a press release issued by An Garda Síochána in February 2024, Detective Superintendent Michael Cryan, Garda National Economic Crime Bureau, said:

"A victim will be made believe that they have met their perfect match online, but are in fact being terribly deceived. It becomes a huge invasion of privacy and a breach of trust which can impact your whole view of people and the possibility of romance, but it's important to remember that this is fraud and it is a crime.

It is often the case that a victim's money is transferred overseas to fund organised criminal activities such as terrorism, human trafficking, people smuggling and even corruption."

Gardaí urge anyone who believes they may be a victim not to feel embarrassed and to come forward, as every report is treated in confidence.

An Garda Síochána's Advice for Online Dating

Use a reputable dating site and stay on its messaging platform rather than moving to social media or texting too quickly. Run a Google Reverse Image Search on profile photos. Be wary of anyone who asks many questions about you but reveals little about themselves. Never send money or share bank details with someone you have met online. Do not invest money in any opportunity offered by an online connection; always seek independent financial advice. Do not download any apps they request, such as AnyDesk. If you become suspicious, save all correspondence and contact your bank and your local Garda Station immediately.

How to Spot an AI-Generated Profile

There are some telltale signs that an online profile may not be genuine. AI-generated photos often have subtle flaws: mismatched earrings, blurred backgrounds near the hairline, or unusually smooth skin. Profiles with very few photos, no tagged friends, and a recent creation date should raise suspicion. If conversations feel overly polished, never include spontaneous details, or the person is always available to reply instantly at all hours, an AI chatbot may be doing the talking.

The simplest test remains the most effective: if someone you have never met in person asks you for money, for any reason, treat it as a red flag.

Where to Report

Anyone who believes they may be a victim of a romance scam should contact their local Garda Station and their bank immediately. Further information is available on the Garda fraud page.