The cryptocurrency sector is well-known for being prone to hacks, yet the rise of artificial intelligence (AI) has increased the potential for fraud and crypto 0 once were considered relatively unsophisticated crypto scams are now becoming far more convincing, scalable, and dangerous due to the rise of generative AI (genAI), deepfakes, voice-cloning, and automated 1 safe from this common crypto scam: Fake Telegram accounts and bots This scam involves creating fake Telegram groups for influencers, projects, and 2 you join, you'll be prompted to "verify" by clicking a 3 you click, malware is installed on your… 4 — Phantom (@phantom) July 1, 2025 AI-Powered Crypto Scams Increase This year alone has witnessed an alarming amount of AI-powered crypto 5 to data from Chainabuse – TRM Lab’s open-source fraud reporting platform – reports of genAI-enabled scams between May 2024 and April 2025 rose by 456%.
This was compared with the same time period during 2023-2024, which had already seen a 78% increase from the previous years.) October 29, 2025 Jardine elaborated that common examples of AI-powered crypto scams include deepfake videos of trusted public figures promoting fake crypto projects, AI-generated phishing websites, fraudulent automated trading platforms that promise unrealistic returns, and voice cloning used to impersonate company executives or family members. AI-generated advertisements are also being used to trick crypto investors and 6 Johnston, code maintainer at Morpheus – a peer-to-peer network for general purpose AI – told Cryptonews that in Spain several people recently orchestrated a crypto scam using AI-generated ads with fake celebrity endorsements.
“Spanish police arrested six people behind the scam that targeted over 200 victims and defrauded them of roughly 19 million EUR,” Johnston 7 Smart, chief of intelligence at Crystal Intelligence, told Cryptonews that the frequency of these AI-powered scams are alarming and growing fast. “AI has made sophisticated scamming accessible to anyone, and you no longer need to be a technical expert to run a convincing operation,” Smart 8 added that Crystal Intelligence published a report about an AI deepfake that used Elon Musk on a Youtube livestream . “The scam took place last year and generated $10,000 overall, but looking at its links on-chain, the operators with other scams may have made orders of magnitude 9 often share payment infrastructure, and we were able to track payments of $1 million related to this campaign,” Smart remarked. 226,000 people are currently being scammed on YouTube by an AI Elon Musk – asking for their 10 account looks like a legit @Tesla account, but it is clearly not, as the username is "tesla.
elon-US" They are using an AI generated Elon talking about the robotaxi and… 11 — Zoli (@zolihonig) October 11, 2024 Web3 Agent Prompt-Injection Attacks Rise Additionally, the crypto ecosystem is witnessing a rise in prompt injection 12 is a security vulnerability where an attacker exploits an AI-agent or a large language model (LLM) to perform unintended and malicious actions. “Prompt injection attacks are a novel technique where an attacker creates inputs that appear legitimate but are designed to cause unintended behavior in machine learning models,” Smart explained “Basically, convincing your ChatGPT to do something that it shouldn’t.” Smart added that prompt-injection attacks have become more concerning as LLMs become more connected to other services such as crypto wallets, internet browsers, email services, and social 13 example, Smart mentioned that while many people are excited to use AI enhanced web browsers – like Perplexity’s Comet – a skilled attacker may craft a prompt injection to embed fake memor i es into the 14 turn, this can send a conversation history to an attacker-controlled email.
A blog post recently published by Brave Browser notes that AI-powered browsers that can take actions on a user’s behalf are powerful, yet risky.
Story Tags
Latest news and analysis from cryptonews
