A deadly combination: Artificial intelligence, crypto, and cyber scammers
Artificial intelligence did not invent fraud, but according to researchers at a leading cyber investigation firm, scams have become faster, more sophisticated, and cheaper. Does the era of deepfakes and fake representatives leave crypto users defenseless?
According to a new report by the TRM Labs intelligence and network analysis lab, the use of artificial intelligence in crypto-related crime has increased by approximately 40% within a year. TRM Labs is a blockchain intelligence and analysis company that provides tools for investigations, detection of suspicious activity, and regulatory compliance for companies and public entities.
In its 2026 index, the overall use of artificial intelligence tools in crypto crime reached a score of 54 out of 100, compared to about 28 in 2024. The fraud sector is already defined by them as mature in terms of artificial intelligence adoption.
The meaning is simple and troubling: artificial intelligence did not invent a new type of crime, but it has made old actions faster, cheaper, and easier to execute. Ari Redbord, a senior executive at TRM Labs, described in a documented interview the change by noting that the technology has lowered the skill threshold required and significantly increased the ability to operate on a large scale.
Instead of an entire team that would need to create fake identities, write messages, and approach victims, a large part of the work can now be automated. According to the report, the share of reported crypto scams involving artificial intelligence tools, such as deepfake videos or chatbots, has grown up to 13 times since 2022. Reported losses from deepfake scams in the last eight months have already exceeded by 263% the total losses recorded in the entire year of 2025.
Even without artificial intelligence, malicious actors have managed for years to steal digital currencies and commit fraud. The big change is that AI can serve as a force multiplier: analyzing weaknesses, creating convincing phishing messages, imitating voices and faces, and carrying out fraud attempts on a scale that was previously not economically viable.
TRM reported that in the first half of 2026, 201 hacks related to digital assets were recorded, a record figure, with a large portion of the losses caused by a relatively small number of major incidents that involved the theft of access keys or login details. The report also points to the growing use of AI in social engineering and the search for security vulnerabilities.
So what can be done? Precisely in an era where a video, voice, or message can be convincingly faked, the most important rule is not to act under pressure. Do not provide passwords, verification codes, a private key, or a recovery phrase to anyone who contacts you, even if they present themselves as a representative of an exchange, a digital wallet, or a security company.
It is not advisable to answer calls from such services; Binance or a crypto wallet will never actually call you. If there is any suspicion, it is advisable to enter the official website or application independently. The combination of crypto and artificial intelligence is not dangerous in itself; both technologies can serve legitimate purposes. The danger begins when the same capability that allows actions to be performed quickly and on a massive scale also falls into the hands of scammers. As forgery becomes more convincing, the public's best defense may well be just one moment of doubt.
