Trust Wallet CEO: Banks Unready for Autonomous AI Attacks

In this article4 sections
Trust Wallet chief executive Felix Fan said on November 18, 2026 that most traditional banks and large corporations are not equipped to face cyberattacks carried out by autonomous AI systems, pointing to research from Google on attackers automating the search for software flaws. Crypto Briefing reported that Fan made the argument in a post on X, tying his concern to work by Google’s Threat Intelligence Group.
Fan’s core point, according to Crypto Briefing, is one of tempo: bank IT teams tend to react to threats after they surface, while language models can scan for weaknesses and act on what they find without waiting for a person to approve each step. Crypto.news reported that Fan wrote he is “paranoid” that most companies are unprepared, singling out banks and corporations that hold large amounts of customer money while their IT departments still work normal weekday hours.
Key facts
- Fan’s November 18 post warned that traditional banks and corporations are not ready for autonomous AI attacks, and he added that he would rather keep funds in Trust Wallet than in banks.
- According to crypto.news, Google’s November 16 AI Threat Tracker from the Google Threat Intelligence Group observed threat actors using large language model frameworks to automate vulnerability discovery and exploitation without human intervention for the first time.
- Google tracked a threat actor between late August and early October using an automated framework to scan organisations across Asia for vulnerabilities, crypto.news reported.
- India’s CERT-In warned on November 4 of rising AI-assisted attacks on fintech firms, digital lenders, payment providers and crypto exchanges, linking some intrusions to an actor designated CRAFTY FALCON that timed operations for weekends and holidays.
- A December 2025 supply-chain attack on Trust Wallet’s browser extension caused losses reported at approximately $7 million by crypto.news and between $7 million and $8.5 million by Crypto Briefing.
Why Fan is weighing in
Fan is not a neutral observer. He has consistently promoted self-custodial wallets, in which users hold their own keys instead of trusting a bank or exchange, as a safer option as AI threats grow. Crypto Briefing reported that Trust Wallet says it has more than 220 million users, giving Fan both an audience and a commercial interest in the argument that centralised custody is the weak link. He became CEO in February 2026, succeeding Eowyn Chen, and in October 2026 announced new operating procedures and security measures aimed at hardening the wallet against AI-driven attacks.
The wallet’s own record is mixed. Besides the December 2025 extension compromise, crypto.news reported that Changpeng Zhao, the Binance co-founder, warned users about a fraudulent Trust Wallet website that tried to get visitors to connect their wallets, and that on November 6 Trust Wallet advised iPhone users to install the latest software updates after reports of malware aimed at older operating systems. Because users control their private keys, the company cannot freeze or recover funds on its own.
Why it matters
The warning lands on a structural tension in traditional finance: large institutions run on layered approvals, legacy systems and change-management processes built for a slower era of threats, while automated attackers do not keep office hours. CERT-In’s findings that intrusions were scheduled for weekends and holidays, when monitoring is thinner, make the staffing gap Fan describes concrete rather than theoretical. For crypto users, the counterpoint is that self-custody shifts the burden of security onto the individual, and holdings are not covered by deposit insurance the way eligible bank deposits are in the United States. The wider industry has its own exposure: crypto.news reported that Binance disclosed attackers compromised an internal AI tool during the events leading to its $6.34 billion breach on October 29.
What to watch
John Hultquist, chief analyst at Google’s Threat Intelligence Group, said attackers are still going after familiar software weaknesses but that defenders will have to embrace agentic defense or risk falling behind, per crypto.news. Google said it is developing its AI Threat Defense offering using Gemini, security technology from Wiz, automated patching through CodeMender and expertise from Mandiant. How quickly banks and exchanges adopt continuous monitoring and faster patching, as CERT-In recommended, is the measurable test of Fan’s claim. Nothing in the November 18 post tied the warning to any specific token or price move.
Disclaimer: This article is for information only and is not investment, financial or trading advice. Cryptocurrency prices are highly volatile. Always do your own research.
Sources: Crypto Briefing, crypto.news


