verifylabs logo
About Us FAQs Pricing Blog Sign up or Login Detected a Deepfake?
About Us Use cases by sector Pricing Blog
Sign up or Login Detected a Deepfake?

There was a time, not so long ago, when a familiar voice on the phone settled the matter. You knew your finance director’s laugh, your mortgage broker’s slightly weary “right then”, the particular way your chief executive clears her throat before saying something expensive. Recognition was a kind of security system, which served us well since the invention of the telephone.

That system has quietly been switched off. Most businesses simply haven’t noticed yet.

The voice you trust, annexed by someone else

Consider a couple in West Michigan. This month, Brian and Wendy VanDoeselaar lost their entire closing fund to scammers who combined AI voice cloning, spoofed email and spoofed caller ID to impersonate the mortgage professional they had been working with for weeks. The sum was $66,000. What undid them wasn’t carelessness. He heard a familiar voice, on a number that matched his lender’s, confirming wiring instructions that arrived in an email that looked right; three layers of fakery reinforced one another before any single one raised a flag. 

If you’re tempted to think this only happens to ordinary people buying ordinary houses, look at who was targeted in August. A coordinated voice-phishing campaign went after several of Wall Street’s largest asset managers, using cloned voices to mimic the tone and phrasing of executives and colleagues in order to coax staff into surrendering credentials or granting system access. Two Sigma, Citadel, Point72 and Millennium Management all faced attempted breaches; Two Sigma confirmed it stopped the attack with no data compromised. These are firms with security budgets most of us can only dream of. The criminals clearly thought the clones were good enough to be worth the effort. 

Where the money goes afterwards

Fraud is only half the story. Stolen money has to be washed, and deepfakes are increasingly how criminals get through the front door of the financial system in the first place. In February, a Ukrainian national pleaded guilty in the US to running OnlyFake, a site that generated and sold digital fake ID documents; prosecutors said it produced more than 10,000 fraudulent IDs, helping customers slip past Know Your Customer checks at banks and crypto exchanges. In the Netherlands, a case running through the Amsterdam courts involved a man who opened 47 fraudulent bank accounts by defeating a bank’s selfie-versus-ID onboarding check with deepfake and face-swap imagery. 

The scale of what sits behind all this is sobering. INTERPOL’s Operation First Light, which ran from January to April this year, targeted social engineering scams and the money laundering around them, leading to 5,811 arrests and the interception of $293 million across 97 countries. In Thailand, investigators found the digital wallet of one suspect, aged just 20, had processed more than $122.5 million in ten months. In Eswatini, police uncovered a gang that had built a realistic replica of a Brazilian police station, complete with fake uniforms and signage, then posed as Brazil’s Federal Police on video calls to talk victims into moving money for “safekeeping”. 

The carpentry involved sounds almost whimsical, until you realise that the most convincing deepfake scams merge real and synthetic content. AI bad actors are increasingly blending real and fake to make their “kills”.

Why “we’ll train our people to spot it” isn’t enough

The instinct of most organisations is to run an awareness session and hope for the best. The evidence is not kind to that plan. Veriff’s 2026 research, based on a survey of 3,000 people across the US, UK and Brazil, found that people are barely better than a coin toss at telling a real video or image from a fabricated one, with video the hardest format of all. In Britain specifically, 22% of people say they never verify suspicious content, the highest rate of any country surveyed. 

Meanwhile the machinery of fraud is being industrialised. Entrust’s 2026 Identity Fraud Report, drawn from more than a billion identity-verification events, found that deepfakes now account for one in five biometric fraud attempts. Here at home, UK Finance reported payment fraud losses of £1.28 billion in 2025, with authorised push payment fraud up 19% to £576.4 million, and said organised criminal groups are increasingly using deepfakes, cloned voices and synthetic identities to impersonate trusted people and bypass identity checks. 

The law has noticed, even if the boardroom hasn’t

This is where “optional extra” finally stops being a defensible position. In the UK, the failure to prevent fraud offence came into force on 1 September 2025 for large organisations, and a company can now be criminally liable for fraud committed for its benefit, including deepfake-enabled fraud, unless it can show reasonable prevention procedures were in place. Across the Channel, Article 50 of the EU AI Act has applied since 2 August 2026, and UK businesses serving EU customers fall within its scope. 

Put plainly: when a deepfake gets through, the question will no longer be “how were you to know?” It will be “what did you have in place?”

What changes when you can check

None of this means abandoning a good process. Call-backs on known numbers, second sign-offs and a healthy scepticism about urgency all still matter. But process relies on someone deciding to be suspicious, and the whole art of a good deepfake is that it gives them no reason to be.

That is the gap a detector fills. VerifyLabs.AI was independently benchmarked in 2026 at 98% accuracy, runs without any API integration, and works across iOS, Android, desktop and Chrome, so the check can happen on the call, in the inbox or at onboarding, wherever the doubt arises. It won’t replace judgement. It gives judgement something solid to stand on again.

The familiar voice used to be the proof. Now it’s the question.

Sources

  1. Resemble AI, “The Deepfake Watchlist: Week of September 11–17, 2026” (17 Sep 2026), summarising CNN, “How two homebuyers in Michigan lost $66,000 in a suspected voice-cloning scam” (15 Sep 2026). resemble.ai/resources/the-deepfake-watchlist-week-of-september-11-17-2026
  2. Resemble AI, “The Deepfake Watchlist: Week of July 31 – August 6, 2026” (7 Aug 2026), summarising Investing.com/Bloomberg, “Wall Street hit by wave of AI-powered ‘vishing’ cyberattacks”. resemble.ai/resources/the-deepfake-watchlist-week-of-july-31—august-6-2026
  3. INTERPOL, “Over 5,800 arrests, USD 293 million intercepted in global fraud bust” (9 Jul 2026). interpol.int
  4. Adaptive Security, “11 Deepfake Attack Examples: Real-World AI Fraud Cases” (26 Jun 2026), citing a February 2026 US Department of Justice announcement. adaptivesecurity.com/blog/11-deepfake-attack-examples-2026
  5. Brightside AI, “Deepfake Fraud in 2026: The $3.7B Problem and How to Defend” (19 Jul 2026). brside.com/blog/deepfake-fraud-losses-2026
  6. Veriff, “What deepfake fraud actually costs businesses in 2025–2026” (17 Jul 2026). veriff.com/fraud/deepfake-fraud-cost-2026
  7. Veriff, “Deepfakes Report 2026 UK” (May 2026). veriff.com/resources/ebooks/deepfakes-report-2026-uk
  8. DeepStrike, “Deepfake Statistics 2026” (Aug 2026), citing the Entrust 2026 Identity Fraud Report. deepstrike.io/blog/deepfake-statistics-2025
  9. IBTimes UK, “UK Fraud Losses Hit £1.28B as Banks Warn AI Makes Scams Harder to Spot” (12 Jul 2026), reporting UK Finance’s Fraud Report 2026. ibtimes.co.uk/ai-driven-fraud-costs-uk-1-28-billion-1808188
  10. Internet Safety Statistics, “Deepfake Technology Threats: A 2026 UK Business Guide” (Sep 2026). internetsafetystatistics.com/deepfake-technology-threats

 

Unmasking machines: how VerifyLabs detects AI-generated content

Artificial intelligence can sometimes feel like magic, right? Whether’s its black magic or the good type depends on who’s using it. At VerifyLabs, we believe that understanding the basics of this technology builds understanding of the power it brings. So, let’s explore how AI detection, particularly with VerifyLabs.AI, helps us differentiate between what’s human-made and what’s a Bot Special.

AI: not a human mind but a sharp pattern spotter

It’s helpful to remember that AI doesn’t have a human brain. It doesn’t “think” or “understand” in the way we do. Instead, it’s incredibly good at pattern recognition. It’s like a super-smart detective that can find tiny clues in data that humans wouldn’t see.

Imagine teaching a child to recognise a cat. You show them thousands of pictures of cats – big ones, small ones (ooh kittens!), fluffy ones, short-haired ones, cats in different poses, different lighting. Eventually, the child learns to identify the common features that make a “cat” a “cat.” AI learning works similarly, but on a much grander, faster scale.

Training AI: differentiating between human and synthetic

For VerifyLabs to detect AI-generated content, our AI models undergo extensive “training.” This involves feeding them enormous amounts of data that are clearly labelled:

During this training, the AI doesn’t just look at the surface. It dives deep, analysing countless tiny features and patterns.

What does AI “look for” in content?

The specific “clues” an AI detector looks for vary depending on the type of media (image, video, audio, text), but here are some common principles and examples:

In text:

In images and videos:

In audio:

Classifiers and confidence scores

At its heart, VerifyLabs.AI uses what are called “classifiers” – these are the AI models that have been trained to distinguish between humans and AI. When you upload content to VerifyLabsAI, our system performs the following steps:

  1. Analysis: the content is broken down and meticulously analysed for all the subtle patterns and clues mentioned above.
  2. Comparison: these patterns are then compared against the vast knowledge base the AI gained during its training.
  3. Confidence score: the AI then assigns a “confidence score” – essentially, how sure it is that the content belongs to the “human” category or the “AI-generated” category.
  4. Clear result: VerifyLabs.AI translates findings into our easy-to-understand “green circle” (human), “red square” (AI-generated) or “grey bar” (more investigation advised) indicator. We believe this clear, simple visual helps you make informed decisions without needing to be an AI expert yourself.

The arms race: why constant innovation is crucial

AI technology is advancing at lightning speed. This means that just as AI models get better at creating deepfakes and synthetic content, our AI detection models must also evolve to keep pace. VerifyLabs.AI is part of a highly-skilled team dedicated to continuous research and development.

Our goal is to be your trusted partner in the age of AI. By understanding a little more about how AI detection works, you can appreciate the sophisticated technology working behind the scenes to show you what’s real and protect your digital peace of mind.

Trust and Verify.

The “deepfake dilemma”: understanding the threat and how VerifyLabsAI protects you

Welcome, digital citizens! In a world where our lives are increasingly online, it’s more important than ever to know what’s real and what’s not. 

You’ve probably heard the term “deepfake” floating around—perhaps in a news story, a viral video, or a cautionary tale. But what exactly are deepfakes, why are they such a big deal and most importantly, how can you protect yourself and your loved ones from their deceptive power? At VerifyLabs, we’re shining a light on this growing challenge and giving you the tools to navigate the digital landscape safely.

What exactly is a deepfake? It’s more than just a photoshopped image!

Think of deepfakes as super-advanced, AI-powered fakes. Unlike a simple Photoshopped image, which manipulates pixels, deepfakes use sophisticated artificial intelligence (AI) and machine learning to create entirely new, realistic-looking images, videos, or audio clips. They can make it appear as though someone said or did something they never did, often with alarming realism.

The “deep” in deepfake comes from “deep learning,” a branch of AI that uses neural networks to learn from vast amounts of data. In the case of deepfakes, an AI model might be fed thousands of images or hours of audio of a person. It then “learns” their facial expressions, voice patterns, and mannerisms so well that it can generate new content featuring that person doing or saying anything the creator desires. Scary, right?

Why are deepfakes such a big deal in 2025?

The deepfake landscape has evolved dramatically. In 2023, there were around 500,000 deepfakes shared. Fast forward to 2025, and projections suggest that this number could skyrocket to eight million. That’s a huge jump, and it tells us a few important things:

These advancements mean deepfakes are no longer just a novelty or a niche concern. They’re a mainstream tool, easily accessible to both sophisticated criminals and opportunistic bad actors.

The growing threat: where deepfakes cause trouble

Deepfakes are popping up in various unsettling ways, impacting individuals, businesses and even our society at large.

Protecting yourself and your loved ones: practical steps

While the landscape can seem daunting, there are practical steps you can take to become a more discerning digital consumer and protect yourself:

  1. Be skeptical: if a video, audio clip, or image seems too good to be true, too shocking, or out of character for the person depicted, pause and question it. A healthy dose of skepticism is your first line of defence.
  2. Verify the source: Before you share anything, especially controversial or sensational content, check where it came from. Is it from a reputable news organisation? An official social media account? Or is it from an unknown or suspicious source? Be wary of content that suddenly appears out of nowhere without context.
  3. Cross-reference information: if you see something concerning, try to find reliable sources reporting the same information. If only one obscure source is sharing it, that’s a red flag. Look for confirmation from mainstream media, official government channels, or trusted experts.
  4. Look for inconsistencies (harder now: older deepfakes often had tell-tale signs: poor lip-syncing, unnatural blinking, inconsistent lighting, or odd movements. While newer deepfakes are much better, sometimes subtle glitches can still appear. Pay attention to:
    • Unnatural facial movements: do expressions seem off or stiff?
    • Poor lip synchronisation: do the words match the mouth movements?
    • Inconsistent lighting or shadows: does the lighting on the person match the background?
    • Odd blinks or eye movements: do they blink unnaturally or too little?
    • Blurry edges or distortions: look for subtle anomalies around the person’s outline or in the background.
  5. Secure your digital footprint: the less material available online that can be used to train deepfake models, the better. Review your privacy settings on social media. Be mindful of what photos and videos you share publicly. Consider limiting access to your old content.
  6. Use verification tools: this is where VerifyLabs.AI comes in! Instead of relying solely on your eyes and ears, powerful AI-driven tools like our deepfake detector are designed to analyse digital media for signs of manipulation. Our app and browser extension provide a quick and easy way to get a clear answer on whether content is human or AI-generated.

VerifyLabs.AI: trust, but Verify

At VerifyLabs.AI, we believe that everyone deserves to feel safe and confident in the digital world. That’s why we’ve developed an intuitive iOS app that puts sophisticated AI detection technology right in your pocket. With our clear “green circle” for human and “red square” for AI-generated content, we make it simple for you to verify images, videos, audio, and text in moments.

As deepfakes continue to evolve, so too will our technology. Stay informed, stay vigilant, and always verify first.

verifylabs logo
© VerifyLabs.AI 2026. All rights reserved.