AI deepfake scams are becoming a more practical risk for UK shoppers, families and business leaders in 2026. Criminals are using fake voices, videos and images to make requests look and sound real — sometimes to push urgent payments, impersonate relatives, or impersonate senior staff. UK government material now treats deepfakes as a fast-growing fraud risk, and its 2026 fraud strategy says criminals are already using them to trick people into handing over money and to impersonate loved ones. (gov.uk)
What has happened
Deepfakes are no longer just a novelty. In 2026, the UK Government said deepfake threats are rapidly evolving, with criminals using fake audio, video and images to support scams and abuse. It also said the scale is rising quickly, citing around eight million deepfakes shared in 2025, compared with half a million two years earlier. (gov.uk)
Government guidance for the public now explicitly warns that fraudsters may use AI to create deepfakes of family, friends, bosses and even trusted organisations. It also highlights a common pattern: authority, urgency and pressure to act quickly. (stopthinkfraud.campaign.gov.uk)
This matters because scammers do not need to break into your accounts if they can persuade you to authorise a payment, share a code, or hand over login details yourself. The Fraud Strategy 2026 to 2029 says the government sees deepfakes as a real vulnerability in consumer and business account security. (gov.uk)
How UK shoppers are being targeted
For shoppers, deepfakes can appear in fake adverts, fake influencer endorsements, counterfeit marketplace listings and “too good to be true” offers. The FCA has warned about online scams that use fake celebrity endorsements and imagery to lure people into frauds, and the government has recently issued guidance to help second-hand shoppers avoid counterfeit goods. (fca.org.uk)
Deepfake tools can also make a scammer sound convincing on a call or in a voice note. A message that appears to come from a familiar brand, a courier, or a bank may be fake, even if the voice sounds reassuring. (stopthinkfraud.campaign.gov.uk)
How families are being targeted
Families are being targeted through “grandparent” style scams, emergency messages, and voice-clone fraud. The UK Government says criminals can use deepfakes to impersonate loved ones, and the NCA has said voice cloning can help deceive victims into believing they are speaking to a colleague, friend or family member. (gov.uk)
A common trick is emotional pressure: “I’m in trouble”, “I need money now”, or “don’t tell anyone”. In a busy moment, that can override normal caution. Government advice says to be especially suspicious of false urgency. (stopthinkfraud.campaign.gov.uk)
How bosses and staff are being targeted
For businesses, one of the most worrying uses of deepfakes is impersonating a senior manager or finance lead. A fake video call or voice message can be used to approve a transfer, change payment details, or ask payroll for sensitive information. The government’s 2026 fraud strategy says criminals are using deepfakes and related social engineering tactics to gain access to consumer and business accounts. (gov.uk)
This is especially relevant for teams that rely on remote working, WhatsApp, SMS, email or quick approval loops. The less verification there is, the easier it is for a scammer to slip through. (stopthinkfraud.campaign.gov.uk)
Why it matters now
The immediate danger is not just technical fraud. It is trust fraud. Deepfakes exploit the fact that people trust what they hear and see, especially when it appears to come from someone they know or an institution they recognise. The UK Government says deepfakes are making it harder for people to know what content is real and are increasing the risk of scams. (gov.uk)
There is also a wider safety issue. In January 2026, the UK Government made clear that non-consensual deepfake intimate images are illegal, reflecting how quickly abusive uses of the technology have become a public concern. (gov.uk)
Warning signs to watch for
- Urgency: you are told to act immediately, with no time to check.
- Secrecy: you are asked not to tell anyone else.
- Pressure to pay: the request is for a bank transfer, gift card, crypto payment or other hard-to-recover method.
- Odd verification habits: the caller discourages you from ringing back or checking through a known number.
- Subtle mismatches: the voice, wording, timing or tone feels slightly off.
- Unexpected contact: a supposed bank, boss or relative gets in touch in a way they normally would not. (stopthinkfraud.campaign.gov.uk)
What readers should do now
If you think you may have been contacted by a deepfake scam, stop and verify before doing anything else. Government advice recommends treating unexpected contact with caution and using known, trusted contact details rather than the number, link or account in the message itself. (gov.uk)
- Pause the conversation. Do not let urgency force a decision.
- Call back on a trusted number. Use one you already know, not the one in the message.
- Ask for a second check. For family, use a pre-agreed code word if you have one.
- Verify payment changes separately. In business, confirm by a second channel and ideally with two people.
- Report it quickly. If money has been sent, contact your bank at once and report fraud through the proper channels. Government guidance points people to Action Fraud for reporting. (gov.uk)
How to stay safer next time
Deepfake scams work best when people rely on one channel only. The safest approach is to build routine checks into everyday life.
- Use a family code word for emergencies.
- Set payment verification rules at work for bank detail changes and urgent transfers.
- Limit public audio and video where possible, because scammers may use social content to mimic voices and mannerisms. This is a practical precaution rather than a guarantee. (stopthinkfraud.campaign.gov.uk)
- Keep software and account security up to date, including strong passwords and multi-factor authentication where available.
- Teach older relatives and teenagers the basics of deepfake fraud, because both groups may be targeted in different ways. This is an inference from the scam patterns described by official guidance. (gov.uk)
What businesses should do
For employers, the practical lesson is to make it harder for anyone to authorise money or data changes on a single message alone. The government’s fraud strategy says customer authentication and account security are vital in cutting off these routes. (gov.uk)
- Require out-of-band verification for urgent finance requests.
- Use call-backs to pre-agreed numbers before changing supplier bank details.
- Train staff to question “CEO urgency” and similar pressure tactics.
- Limit who can approve payments and keep an audit trail.
- Make it normal to challenge unusual requests, even from senior staff.
Where official guidance is heading
The UK Government says it will work with the private sector to develop consumer advice on fraud, and it is also pursuing broader anti-fraud measures through its 2026 to 2029 strategy. It has separately launched deepfake-related work, including a Deepfake Detection Challenge in January 2026, which shows the issue is being treated as an active policy problem rather than a niche tech issue. (gov.uk)
That said, detection technology is still described by government material as early-stage and a developing market, so prevention and verification remain more reliable than trying to spot every fake by eye or ear. (gov.uk)
Key takeaway
Deepfake scams in 2026 are less about perfect technology and more about convincing people to act fast. The best defence is to slow down, verify through a trusted channel, and put simple family and workplace checks in place now. (stopthinkfraud.campaign.gov.uk)