Philosophers and futurists have likely made similar declarations at many other (or every other) milestone in human history, but it does feel like as 2026 draws to a close, the world is now changing more quickly than ever before.
The pace at which artificial intelligence capabilities are expanding and improving only continues to accelerate — often exponentially, and seemingly overnight. How to prepare for, predict, and not only survive but also thrive in a future that feels as uncertain and foreign as this one is, for many, the source of great anxiety.
That uncertainty and anxiety extend across all industries and verticals, including the fraud and financial crime space. In an attempt to prepare for the months ahead, we’ve assembled a group of both internal and external experts from all around the world to consider how we might expect fraud and financial crime to continue to evolve in 2027.
—
Diego Baldin
BioCatch global advisory director for Brazil
Location: Brazil
“In 2027, proving who is behind a transaction will no longer be enough. As authentication gets stronger, criminals will increasingly manipulate legitimate customers into moving money themselves, making intent, device context, and the trustworthiness of the destination just as important as identity. Fraud prevention will therefore need to connect behavioral intelligence, device intelligence and cross-institutional network signals to answer not only ‘is this really you?’ but also ‘why are you doing this, from what environment, and where is the money going?’"
Sharell Barshishat
BioCatch global advisory director for North America
Location: U.S.
“Banks will keep getting hit by large-scale attacks that use AI to adapt and change in days, hours, and even minutes to get around controls. Refined anomaly detection will help, but banks will have to get faster in their responses, which will be hard given the change-management and governance processes they have to follow. We will have to reckon with a faster enemy than ever before. The goal of these attacks will be to create accounts that are sold for profit and/or used to launder money. Regulators will keep issuing large fines, as it becomes more and more evident that banks can do more about mules.
"Large banks will take a full block posture on agentic traffic because of liability and security concerns for their customers. They will struggle, though, to identify agents that don’t openly present themselves, such as those running through anti-detect browsers or successfully spoofed by fraudsters. Smaller banks may see an opportunity in accepting this traffic as a way to appeal to a consumer market that wants this level of access from banking.”
Dan Benisty
Founder of DB Compliance Advisory and former head of compliance for Northern Europe at Western Union
Location: France
“In 2027, one of the biggest shifts might be the convergence of fraud and money laundering around the money mule account. Employment and investment scams, increasingly scripted by AI-generated voices and deepfake advisors, will keep feeding mule networks (often made up of young people) recruited on social media and messaging apps. At the same time, the EU's AML Regulation becomes applicable in July 2027 and the new Payment Services Regulation will make banks more liable for impersonation fraud, so a fraud alert, a suspicious activity report, and a refund claim will increasingly be about the same customer. The institutions that come out ahead will be those that stop running fraud and AML/CFT in separate silos and connect behavioral signals, beneficiary-side monitoring, and mule detection in a single risk view.”
Richard Booth
BioCatch SVP of emerging solutions and network
Location: Australia
"Criminals will accelerate the weaponization of AI to increase the volume, velocity, and variety of scams and fraud. This will extend to the herding of mule accounts by compromising scam victims to engage in muleing activities on a larger scale than in previous years. Australian banks are already leading the organizational shift to a more consolidated operational infrastructure, and we will see this continue. Previously separated fraud and financial crime teams will operate closer than ever before, and multi-skilled team members will leverage AI tools to create a single view of risk rather than the siloed views of the past.”
Subhashish Bose
BioCatch global advisory director for APAC
Location: Singapore
“In 2027, fraud risk detection will focus on identifying bad intent in order to stop money from leaving mule accounts. As AI-driven scams and sophisticated mule networks make malicious accounts harder to distinguish from legitimate ones, enhancing receiving-account risk detection models will become critical. Behavioral intelligence combined with real-time cross-bank network signals will become central to disrupting both scams and mule activity.”
Ayleen Charlotte
Founder of S.A.F.E. Academy, consultant in victim-centered fraud response, fraud victim advocate, international speaker
Location: The Netherlands
“In 2027, I believe we will finally start treating fraud as more than a financial or technological problem and pay far greater attention to what it does to the person behind the transaction. As AI makes scams more personalized, convincing, and harder to recognize, the idea of a ‘typical vulnerable victim’ will become obsolete. Financial institutions will need to look beyond detecting and stopping the payment and recognize that how we respond to the person afterwards is part of fraud prevention itself. The institutions that understand the human impact of fraud will not only support victims better [but also] be better equipped to prevent further harm.”
Jonathan Frost
BioCatch global advisory director for EMEA
Location: U.K.
“When an OpenAI agent breached Hugging Face during a test, no one told it to attack anyone. It was given an objective and found its own route. In 2027, the unauthorized fraud risk will be an agent with a fraud objective, while the authorized risk will be customer manipulation with machine speed and accuracy. The defense is knowing whether a human is in the session.”
Josue Martinez
BioCatch global advisory director for Latin America
Location: Mexico
“In 2027, AI-enabled synthetic identities and the misuse of shell companies could shift financial crime risk from individual fraudulent accounts to interconnected networks operating behind seemingly legitimate identities and businesses. Financial institutions’ strategic priority should be to unite fraud prevention and AML around continuous, network-level risk intelligence and not isolated onboarding decisions. Integrating identity and beneficial ownership verification with behavioral, device, and transaction intelligence, network analytics, and lawful information sharing can enable earlier disruption of criminal networks, reduce financial and regulatory exposure, and protect customer trust without compromising the legitimate customer experience.”
Dirk Mayer
Head of anti-fraud consultants at RISK.IDENT GmbH
Location: Germany
“1.) Faster and more personalized: AI optimizes social engineering. [Authorized push payment fraud] will increase dramatically once again.
"2.) The European Union ID becomes a target. Starting January 2, 2027, the European Union ID is set to be introduced — the ideal entry point for fraudsters. They will attempt to harvest customers’ authentication data on a massive scale and use the introduction to manipulate [those customers] directly.
"3.) The PSR is becoming a hot topic and is already influencing liability issues even before it takes effect. Starting with its publication, financial institutions will be held to new standards, and court rulings are even more clearly in favor of consumers.”
Patricia Peck
CEO and founder of Peck Advogados
Location: Brazil
“In 2027, financial institutions will expand identity verification beyond customers to authenticate autonomous and human-controlled digital agents, using continuous, risk-based controls to confirm who — or what — is initiating each transaction. AI will become central to real-time anti-fraud strategies, particularly for detecting synthetic identities and deepfake-enabled social engineering across voice, video, and digital channels. To identify behavioral deviations earlier and strengthen customer security, institutions will process more personal and behavioral data under the principle of exhaustion — using all relevant, proportionate, and lawfully available signals while maintaining transparency, purpose limitation, and robust data governance.”
Susan Potgieter
Banking risk and financial crime thought leader and former chief risk officer at South African Banking Risk Information Centre (SABRIC)
Location: South Africa
“The exponential adoption and exploitation of AI will enable criminals to weaponise trust and data at unprecedented scale while challenging prevention, enforcement and deterrence. AI-enabled impersonation scams will become increasingly convincing and difficult to detect, exploiting customers’ fear of banking crime, limited digital confidence, and trust in reputable organizations to manipulate them into disclosing sensitive information or authorizing fraudulent transactions. Third-party data breaches will remain a critical enabler of targeted scams, exposing vast volumes of personal information that criminals will use to build detailed victim profiles and execute highly personalized attacks at scale. The convergence of cybercrime and cyber-enabled fraud will intensify pressure on existing legal frameworks, compelling stronger cross-border cooperation, more effective enforcement and greater use of legislative powers to strengthen deterrence and hold perpetrators accountable.”
Kevin Purkiss
Fraud and scam strategy advisor, and former vice president of fraud management at Royal Bank of Canada
Location: Canada
“In 2027, financial institutions will face growing expectations to demonstrate how effectively they protect customers from scams. As criminals exploit AI and increasingly sophisticated social engineering, threats will continue to evolve faster than regulation and traditional fraud controls can respond. Meeting this challenge will require institutions to move beyond detecting suspicious activity [and] toward proactive prevention, combining better intelligence, timely intervention, and behavioral training that equips fraud agents to recognize manipulation and engage effectively with customers. Ultimately, success will be measured not simply by the fraud we detect, but by the harm we prevent.”
Jason Richards
Former special agent with the FBI
Location: Thailand
“Today, the most significant threat in the global fraud landscape is the proliferation of scam compounds, which resemble the criminal enterprise depicted in New Jack City, where Wesley Snipes’ character takes over an entire apartment complex and turns it into a self-contained criminal operation, controlling the workforce, infrastructure, security, and profits. In 2027, I predict these labor-intensive operations will increasingly evolve into AI-driven, highly automated criminal enterprises capable of conducting fraud on a much larger scale with significantly fewer people. While this shift may reduce the demand for trafficking victims, it will also make intelligence gathering and evidence collection increasingly difficult as fewer individuals have firsthand knowledge of these operations. Meanwhile, rising nationalism and geopolitical tensions will erode international law enforcement cooperation, allowing these increasingly decentralized and technologically sophisticated criminal organizations to operate with greater impunity.”
Alex Robinson
BioCatch global advisory director for EMEA
Location: U.K.
“I expect one or more countries or regions to experience large-scale automated fraud attacks across multiple banks, with AI scaling and adapting the attack method quicker than banks using rules-based fraud systems can respond. The pace of attack adaptation will lead to high volumes of impacted customers, media attention, and questions about which banks can be trusted to protect their customers. Banks need to adapt defense methods to continuous real-time behavioral profiling to be able to respond to evolving attack typologies and adapt response methods to crime gangs using sophisticated AI methods.”
Nate Schneemann
BioCatch global advisory director for North America
Location: U.S.
“I expect banks to remain cautious about allowing AI to transact on behalf of their customers as they work to understand the risks, opportunities, and controls needed to do so safely. Fraudsters will continue leveraging increasingly sophisticated AI capabilities and open-source technologies to enable individual actors to operate at a level previously reserved for organized criminal groups, while adversarial groups scale attacks that put enormous pressure on banks and their financial crime prevention infrastructure. I’m optimistic that financial institutions in an environment of increasing regulatory pressure will recognize trust as a driver of growth and help build the new risk prevention infrastructure necessary to harness the promise of AI as the digital world is reinvented.”
Frank Schwab
Independent board director and strategic advisor, former CEO of Fidor Solutions and GIZS (paydirekt), and former director of strategy and innovation at Deutsche Bank
Location: Germany
“In 2027, I expect AI to make personalized scams cheaper, more convincing, and easier to scale. Leading banks will intervene earlier against customer manipulation, rather than rely on identity checks alone. Fraud and anti-money-laundering teams will work more closely to disrupt criminal networks and not just flag individual transactions. And boards will start to measure success by reduced customer harm and disrupted criminal money flows — not more alerts.”
Charanjeet Singh
Fraud risk and financial crime advisor at C S Risk Consultancy
Location: India
“Agentic AI is going to disrupt the fraud risk management, from the bad actor side as well as from the counter-fraud side. Those who are simply blocking AI agents will be forced to [let them in], and the challenge [will be] to identify agentic activity and then segregate bad and good agents.”
Albert Stoetzer
Global ML/TF/fraud scenario enhancement and risk coverage specialist at Rabobank, founder of Benelux FinCrime Fighters platform 112Wwft.nl
Location: The Netherlands
“My biggest prediction for 2027 is the rise of increasingly autonomous digital scam farms, deploying hundreds, thousands, or even tens of thousands of malicious AI agents per operation. Working around the clock to manipulate victims into making fraudulent payments, these operations could scale far beyond today’s labor-intensive scam compounds in Southeast Asia. As a fin crime fighter, I am fascinated by the prospect of a hybrid battle, with humans and AI agents working together on both sides, constantly challenging and outsmarting each other. What excites me most is the opportunity to combine human judgement with AI’s speed and scale to disrupt scams earlier and protect more people.”
Nicolas Strömbäck
CEO of NS Consulting, former group MLRO and senior group compliance officer at Ikano Bank
Location: Sweden
“Having been in both the [governance, risk, and compliance] and AI spaces of late, I think that AI governance is a must for firms to counteract illicit activities. Why governance, not strategy? I hear firms saying they are implementing AI, but there is no consideration of how to govern this space in relation to their business. Without such governance, chances are good that a lot of money will be spent on solutions that will not improve the possibilities of finding and reporting illicit activities. Strategy without governance will not yield desired results.”
Donna Turner
Operating partner, CXO Partners’ compliance and fraud prevention practice, former COO at the Zelle Network
Location: U.S.
“The acceptance and application of AI in fighting back against fraud will see a notable increase, recognizing its inherent ability to identify emerging trends and counter the same much more rapidly then the traditional analytical cycle. The criminal use of false signals generated solely to disrupt fraud detection analytics and countermeasures will be more prevalent, which will necessitate the convergence of traditional analytics and AI analytics to maintain accuracy and avoid under disruption. The broader use of AI agents will allow for deeper penetration of high false positive vectors like check and deposit fraud, which in turn will pave the wave for future use cases. Additional points of verification will be deployed to counter the ongoing growth in dispute volumes, targeting both unauthorized and first party fraud disputes.”
Tomislav Vazdar
Founder and CEO of Riskoria
Location: Croatia
“In 2027, scam operations will hand the volume work to AI (first contact, screening victims, and keeping conversations going in any language), while human scammers focus on the people most likely to pay. This is already starting, and next year it becomes standard practice. For banks, it means more scams that look like normal customer behavior because the customer logs in and approves the payment themselves. The institutions that cut losses will be those that can spot a customer paying under someone else’s direction, not just those that verify who logged in.”
Erin West
Founder and CEO of Operation Shamrock
Location: U.S.
“In 2027 the scam compounds will keep moving, and you should expect to see them take root in places most of us aren’t watching yet, from West Africa to the Pacific, because every crackdown in Cambodia or Myanmar teaches the bosses to pack up and rebuild somewhere new. At the same time, AI will let a single trafficked worker carry on dozens of conversations at once, in flawless English, with a face and a voice convincing enough to fool your own family. If your fraud program is counting on the customer to notice something is off, I need to tell you that the customer will have nothing left to notice.”
Katy Worobec
Independent fraud expert, former managing director of economic crime at UK Finance
Location: U.K.
“Purchase scams have long been overshadowed by higher value crimes like investment and romance fraud, yet they affect far more people through constant exposure to fake ads on social media and in gaming apps. Because many victims are reimbursed or don’t bother reporting, the true scale is significantly under-recorded, masking the growing economic harm. Increasingly, scammers emulate well-known brands yet operate from overseas, quietly draining millions from the country’s economy. With AI making fake ads harder to spot, this trend is set to escalate sharply, positioning purchase scams as a defining fraud threat in 2027.”
—
Resources:
- Global survey: The future of digital trust: AI agents and the speed of fraud and financial crime
- Report: 2026 Global Scams
- Analyst research: Five forces disruption global fraud prevention by 2030
- Blog: From Firebase to agentic AI: One prompt away from the scam pipeline endgame
- Blog: Is that your customer, their agent, or someone (or something) else?
