AI in the Service of Scammers: How Crypto Fraud Became a Billion-Dollar Industry

Key points:

• Artificial intelligence is turning crypto fraud into an industry: AI technologies have transformed cryptocurrency scams from labor-intensive operations into automated systems capable of operating at massive scale through automation, personalization, and rapid testing of new schemes.
• In 2025, the volume of illicit cryptocurrency activity reached a record high: total suspicious transaction volume hit $158 billion, nearly 145% higher than the previous year. Fraud schemes alone accounted for about $30 billion, though the real figure is likely significantly higher due to widespread underreporting.
• AI is dramatically increasing the scale, speed, and flexibility of fraudsters: over the past year, AI-enabled fraud grew by roughly 500%. This reflects the rapid integration of generative AI into scam operations — from phishing emails and fake identities to automated money laundering and the creation of synthetic digital personas.
• Automation is creating a strategic imbalance between criminals and defenders: bad actors can launch thousands of small-scale campaigns at minimal cost, while compliance teams and law enforcement must operate within strict legal and procedural constraints.
• Blockchain transparency remains a defensive advantage: AI accelerates fraud but does not make it invisible. All blockchain transactions are permanently recorded and can be analyzed, allowing investigators to identify connections, detect anomalies, and conduct investigations using analytics tools and professional expertise.

Artificial intelligence is reshaping crypto fraud — transforming it from an activity limited by human resources into a large-scale automated system. GetBlock AML Research reviews the industry that cryptocurrency fraud has become in the age of AI.

In 2025, illicit cryptocurrency activity reached an all-time high of $158 billion — nearly 145% higher than the previous year. Of that amount, around $30 billion came from fraud schemes alone, and given the significant underreporting by victims, the real scale is likely much larger.

Over the past year, AI-enabled fraud increased by approximately 500%. The convergence of generative AI, programmable financial infrastructure, and the global accessibility of cryptocurrencies has fundamentally changed the economics, speed, and scale of deception. This report examines how AI is reshaping the entire lifecycle of crypto fraud and what it means for investigators, regulators, and financial institutions.

What AI-powered crypto fraud is — and how it works

AI-powered crypto fraud refers to financial crimes and deception schemes in which artificial intelligence is used to automate, personalize, and scale fraudulent activity. Crypto scams existed long before AI, but AI is changing how they are executed — schemes are launched faster, appear more convincing, and become harder to detect.

How AI scales crypto fraud and illicit activity

Crypto fraud has existed since the earliest days of digital assets. Even large-scale criminal operations once depended heavily on human labor. Investment scams, romance scams, and fake support schemes required prolonged communication with victims. Call centers, operator training, quality control, and shift schedules naturally limited scale.

Today, artificial intelligence removes many of those constraints.

AI does not necessarily create entirely new types of crime. Instead, it amplifies existing ones by reducing reliance on human involvement.

Automating phishing and social engineering

Criminals use generative AI to create convincing phishing emails, fake investment websites, and customer support chatbots within seconds.

Language models can tailor messages to specific individuals or social groups, increasing response rates.

Automated translation tools enable cross-border scam operations without language barriers.

Deepfakes and synthetic identities

AI also accelerates identity spoofing. Synthetic audio and video tools allow criminals to impersonate company executives, romantic partners, or public figures. In crypto, this may appear as fake token endorsements, fraudulent trading signals, or scammers posing as exchange representatives. All of this takes place in an environment where transactions are fast and irreversible — funds can be sent and permanently lost within minutes.

How AI automates money laundering and smart contract attacks

AI-driven fraud is not limited to communication-based deception. Machine learning is used to bulk-test stolen data, optimize transaction routing for laundering, and identify vulnerabilities in digital contracts. As AI tools become more accessible, the barrier to launching sophisticated schemes continues to fall.

What once required large teams can now be automated, standardized, and continuously improved — fundamentally changing the economics of deception.

The key shift is scalability. Human-driven fraud scales with headcount. AI-driven fraud scales with computing power. This creates a more industrialized threat landscape, where automation combines with personalization and cross-platform coordination. Understanding how AI amplifies existing schemes is the first step toward building effective defenses.

Illicit transaction volume in 2025: $158 billion and the industrialization of crypto crime

In 2025, illicit transaction volume reached $158 billion — nearly 145% higher than the previous year. Although the share of illicit activity relative to total transaction volume declined slightly from 1.3% to 1.2%, the absolute increase reflects the expansion of both the ecosystem and criminal capabilities.

By liquidity metrics, illicit entities controlled approximately 2.7% of available cryptocurrency liquidity in 2025. This suggests they are not fringe participants but structurally embedded in the global market.

Fraud growth: $30 billion in 2025 alone

Fraud remains one of the largest threats to everyday users. In 2025, fraud-related activity totaled about $30 billion. However, due to underreporting driven by embarrassment or delayed detection, actual losses may be up to 85% higher than visible data suggests.

Against this backdrop, a sharp behavioral shift is underway: AI-enabled fraud grew by 500% year over year. This is not gradual adoption — it represents a rapid transition to new methods.

Synthetic trust: how AI builds large-scale crypto scam operations>

Trust is the foundation of most successful crypto scams. Investment and romance schemes rely on emotional engagement. Previously, building that trust required prolonged one-on-one interaction.

Generative AI accelerates and multiplies this process.

AI systems can simultaneously conduct hundreds of credible conversations, adapting tone and language to different cultures and audiences. Automated translation removes geographic barriers. The result is “synthetic trust” — communication that feels personal but is algorithmically generated.

How AI accelerates the crypto fraud lifecycle — from phishing to cash-out

AI not only expands reach but compresses timelines. Victim data collection becomes automated. Messages are generated and optimized for maximum engagement. Conversations continue without fatigue or time-zone limitations. The final stage often involves fake platforms that closely mimic legitimate services.

Money laundering is also accelerated through automation. Funds are rapidly routed across multiple digital platforms and instruments, complicating tracking efforts.

For investigators, this means response windows are shrinking.

The economics of crypto fraud: why AI lowers criminal costs

AI fundamentally reshapes cost structures for fraudsters. Humans require training, salaries, and supervision. Once configured, AI systems operate 24/7.

The marginal cost of targeting an additional victim approaches zero.

Scripts and narratives are continuously optimized based on performance data.

New scam variants can be tested rapidly.

This allows small groups to reach scales previously achievable only by large criminal enterprises. Capabilities become more “democratized.”

Why AI creates an imbalance between fraudsters and compliance teams

AI creates asymmetry between attackers and defenders. Criminals can experiment freely — launching thousands of low-cost test campaigns. Losses are minimal; a small percentage of successful attempts is enough to generate profit.

Defenders, however, operate under legal constraints, procedural requirements, and the risk of false positives. Incorrect accusations or account freezes carry serious consequences. Resources are finite.

Closing this gap requires equally scalable AI-powered analytics. Only by matching machine-scale threats with machine-scale detection can the imbalance be reduced.

Blockchain transparency vs. AI fraud: can criminals still be tracked?

Despite increased speed and scale, blockchain remains transparent. All transactions are permanently recorded and available for analysis. Even automated schemes leave digital footprints.

The more automation, the more data. Reused wallets, shared infrastructure, and behavioral similarities create investigative leads. However, technology alone is insufficient. Human expertise remains essential. Accountability decisions require context and coordination between digital platforms and law enforcement.

AI can surface signals within vast datasets, but interpretation and decision-making remain human responsibilities. The future of crypto fraud will be defined by automation, speed, and coordination — and the response must be equally technological and professionally organized.