Logiciel Solutions Contact Us
Success Stories Tech News Investors Contact Us
whitepaper
8 min read

AI in FinTech: KYC, Fraud and Scam Detection.

US consumers reported $15.9 billion in fraud losses to the FTC in 2025, and the costliest schemes now persuade customers to send the money themselves, so the login is genuine and classic fraud controls see nothing wrong.

AI in fintech is how risk teams are responding, from injection attack detection at onboarding to scam scoring before an instant payment is released. The rules moved in 2026 as well: SR 26-2 superseded SR 11-7 and excluded generative and agentic AI, and FinCEN proposed a rewrite of AML program requirements.

The evidence here is reported and estimated losses from US agencies and regulatory texts current to November 2026. It does not measure any single institution's exposure, and true losses are likely higher than the reported figures.

In depth

Fraud Now Moves Faster Than The Controls Built To Stop It.

01

The largest losses now come from customers who are persuaded to pay.

The FTC's Consumer Sentinel Network received about 3 million fraud reports in 2025, with $15.9 billion in reported losses, up from more than $12 billion the year before. Investment scams accounted for $7.9 billion and imposter scams for $3.5 billion, and social media was the costliest contact method at $2.1 billion. The FBI's IC3 logged 1,008,597 complaints, the first year above one million, with about $20.9 billion in reported losses, $7.7 billion of it from adults aged 60 and over. In an authorized push payment scam the customer initiates the transfer, so authentication succeeds and the payment looks routine. A control that only asks whether the right person is logged in will approve every one of these payments.

In shortA control that only asks whether the right person is…
02

Identity verification on its own no longer proves a person is real.

FinCEN's alert FIN-2024-Alert004 warned that criminals use generative AI to alter or fabricate identity documents and selfies to pass onboarding, and it asks institutions to tag related SARs with the key term FIN-2024-DEEPFAKEFRAUD. Gartner predicted in 2024 that by 2026, 30% of enterprises would consider identity verification unreliable in isolation because of deepfakes, and it reported that injection attacks rose 200% in 2023. Synthetic identities, which blend real and fabricated data into a person who does not exist, carried an estimated $35 billion in 2023 losses according to the Federal Reserve's toolkit, and the Boston Fed says generative AI makes them harder to detect. Liveness checks still matter, but on their own they are one layer that attackers already target.

In shortLiveness checks still matter, but on their own they…
03

Instant payment rails leave very little time to act.

The Richmond Fed counted 1,725 banks and credit unions on FedNow in the first quarter of 2026, about 19.7% of US institutions, and listed RTP at 1,193 member institutions. RTP first passed 2 million payments in a single day in February 2026. The UK made APP scam reimbursement mandatory in October 2024, and an independent Frontier Economics evaluation estimated that APP losses over Faster Payments then fell by about 21%. The US has no comparable mandate, though the banking agencies sought public input on payments fraud, including instant payments, in June 2025. Nacha's 2026 risk management rules also extend risk-based monitoring to institutions receiving credits, which makes inbound monitoring a compliance duty. On an instant rail, a scam decision is only useful if it lands before the payment is released.

In shortOn an instant rail, a scam decision is only useful i…
04

The model risk rulebook changed in April 2026.

On April 17, 2026, the OCC, Federal Reserve and FDIC issued revised guidance as SR 26-2, OCC Bulletin 2026-13 and FDIC FIL-15-2026, superseding SR 11-7 and SR 21-8, the 2021 BSA/AML model statement. According to Orrick's analysis, it is most relevant to banks with over $30 billion in assets, requires complexity before something counts as a model, carves out deterministic rules, adds a vendor model section and excludes generative and agentic AI pending a request for information. In the same month FinCEN proposed reforming AML/CFT program rules around effectiveness, with a 12-month implementation period after any final rule, and it remains a proposal. On that definition an ML alert scorer is still a model, while an LLM copilot needs governance from somewhere else. SR 11-7 is gone, but the duty to govern fraud and AML models is not.

In shortthe duty to govern fraud and AML models is not
The detail

Three Questions To Ask Before Any AI Fraud Control Goes Live.

None of this requires a risk committee to choose a vendor first. It requires the committee to know what each model is allowed to decide, how its threshold was priced and who governs the parts that fall outside SR 26-2. Three questions surface all of it, each answerable with one named artifact.

Zone · 01

What may the model decide

Ask whether each use case automates, augments or only advises. Identity checks can auto-approve and send failures to review, sanctions screening keeps a deterministic floor no model can override, and a SAR is always decided and filed by a person. The mode sets the explainability, testing and audit evidence needed later, so it comes before model choice. The artifact is a use case register with a decision mode against each line.

Zone · 02

How was the threshold priced

Ask what the alert threshold costs in total, not how many alerts it removes. Every false positive carries review cost and customer friction, and every miss carries a fraud loss or regulatory exposure. A model that halves alerts but misses one mule network can cost more than it saves. The artifact is a total cost calculation with precision reported at the investigator capacity line.

Zone · 03

Who governs the LLM copilot

Ask where the copilot that drafts alert summaries and SAR narratives sits in the governance framework. SR 26-2 excludes generative and agentic AI, so the copilot needs its own controls: grounding in retrieved case facts, citations to evidence, logged prompts and outputs, and an investigator who edits and approves. The artifact is a copilot control sheet with an audit trail behind it.

By the numbers

The figures that make it a board-level conversation.

$15.9 billion
in fraud losses reported by US consumers to the FTC in 2025, with investment scams alone at $7.9 billion
22,000+
AI-related complaints logged by the FBI for the first time, carrying $893 million in reported losses
$35 billion
estimated synthetic identity fraud losses in 2023
Inside the report

What you'll take away.

01

Step 1 - Assign each fraud and AML use case a decision mode first

Decide whether the system automates, augments or advises before choosing a model, because the mode sets the explainability, testing and audit evidence you will need under BSA, OFAC and consumer protection law.

02

Step 2 - Layer injection attack detection on top of liveness at onboarding

Gartner advises pairing presentation attack detection with injection attack detection and image inspection. Add device intelligence and document forensics, then tag deepfake cases with FinCEN's key term.

03

Step 3 - Score instant payments for scams before the funds are released

Use device, session, payee and behavioral signals such as an active call during a session, and scale the response from a warning to a confirmation question to a hold with an agent call.

04

Step 4 - Govern ML scorers under SR 26-2 and LLM copilots separately

Rate materiality, document validation, fairness tests and reason codes for scoring models, and put copilots under broader risk governance with grounding, citations and logged prompts and outputs.

Questions

Frequently asked.

How is AI used in fintech for fraud and compliance?

AI in fintech verifies identities, scores payments and logins for fraud, prioritizes AML alerts, screens sanctions lists and detects scams in progress. Gradient-boosted models handle real-time scoring, graph models find mule networks and synthetic identity clusters, and LLMs draft alert summaries and SAR narratives.

Humans stay accountable for SAR filings, account closures and adverse decisions, and the program must still satisfy BSA/AML, OFAC and consumer protection law.

Can AI detect deepfakes during KYC onboarding?

Yes, when the checks are layered. Presentation attack detection catches masks and screen replays, and injection attack detection catches virtual cameras and tampered video streams, which Gartner reported rose 200% in 2023.

Document forensics and device signals add further checks. FinCEN's 2024 deepfake alert lists red flags and asks institutions to tag related SARs. No single check is reliable alone.

How do banks use AI to stop scams on real-time payments like FedNow and RTP?

Banks score each instant payment before release using device, session, payee and behavioral signals. Scam models look for signs of coaching, such as an active phone call during a session or a first payment to a crypto exchange.

Responses are tiered, from a warning to a confirmation question or delay to a hold with an agent call. Inbound monitoring catches mule accounts receiving the proceeds.

Has SR 11-7 been replaced?

Yes. On April 17, 2026, the Federal Reserve, OCC and FDIC issued revised model risk management guidance as SR 26-2, OCC Bulletin 2026-13 and FDIC FIL-15-2026. It supersedes SR 11-7 and the 2021 BSA/AML model statement, narrows the definition of a model, carves out deterministic rules and excludes generative and agentic AI. The agencies plan a separate request for information on AI.

How does AI reduce false positives in AML transaction monitoring, and what does compliance cost?

Supervised models re-rank rule-generated alerts using past dispositions and SAR outcomes, so investigators work the riskiest alerts first, and graph features add network context single-account rules lack. The cost pressure is real: LexisNexis Risk Solutions estimated financial crime compliance in the US and Canada at $61 billion a year, with labor the largest component.

Measure success as precision at investigator capacity and total cost, not raw alert count.

What is in the AI in fintech whitepaper?

A 17-minute guide covering the 2025 and 2026 threat data, seven use cases from KYC to SAR copilots, a comparison of model approaches, the regulatory position as of November 2026, explainability and fairness testing, a total cost formula for alert thresholds, a reference architecture, KPIs, a 12-month roadmap and a seven-dimension readiness scorecard for financial crime programs.

How far does the evidence in this whitepaper reach?

It covers reported and estimated losses from US agencies, the Federal Reserve System, Nacha, The Clearing House and the UK Payment Systems Regulator, plus regulatory texts checked through mid-2026. Actual losses are likely higher than the reported figures.

It does not measure any single institution's exposure, vendor names are examples rather than endorsements, and the FinCEN program rule it describes is still a proposal.

Get the whitepaper

Have it emailed to you.

Drop your details and we'll send AI in FinTech: KYC, Fraud and Scam Detection straight to your inbox - no spam, unsubscribe anytime.

Download whitepaper
Next step

Map where your fraud controls decide, before the next scam payment does.

Bring your current onboarding, monitoring and payment fraud stack and the model inventory behind it. We will work through where each control sits in the payment flow, what it is allowed to decide and how it maps to SR 26-2. SECTION 7 - FAQ - 5 to 8 questions

Book a fraud stack review