AI Agents Could Make Crypto Hacks Far Worse
Industry leaders warn that AI agents may turn today’s large crypto hacks into small events by comparison. At the Wyoming Blockchain Symposium 2026, experts discussed both the benefits and dangers of these tools. AI agents can simplify crypto and financial services. Yet they also give bad actors powerful new capabilities.
AI Agents Lower Barriers for Attackers
Ryan Kirkley, CEO of Global Settlement Network, said people wrongly assume agents are always helpful. He noted that AI agents can target Wi-Fi networks, passwords, and wallets at a scale once impossible.
“We think these bridge hacks are bad,” Kirkley said. “It’s pennies. It’s nothing.”
In the past, attacking a person with $20,000 rarely made sense. Now one agent can go after many people at once. Bill Laboon of the Web3 Foundation agreed. The same tools that help honest users also help criminals by cutting friction for everyone.
Clear Limits on Agent Power Are Needed
Panelists debated how much control users should give AI agents. Fahmi Syed of the Midnight Foundation stressed the need for strict parameters. Giving one agent full access to credit cards, Social Security numbers, and all accounts is risky without clear boundaries.
Kirkley called authority limits one of the easier problems to solve. The bigger issue is securing the systems themselves. An agent that is taken over could drain an entire wallet. Laboon also warned about privacy. Even private blockchains can leak metadata. AI tools may later combine that data and expose information users thought was safe.
Trust and Hallucinations Block Wider Use
Most experts agreed that trust is the biggest barrier to adoption. Laboon pointed to the common problem of AI hallucinations.
“My LLMs still occasionally hallucinate,” he said. “I wouldn’t want to put my 401k in the hands of that.”
Richard Shorten of Silvermine Capital Advisors added that the technology has moved faster than public understanding. People need to feel safe before they let agents handle parts of their finances or business.
Legal Liability Creates Major Uncertainty
Kirkley highlighted another key challenge: unclear legal responsibility. When an autonomous agent makes a costly mistake or breaks the law, it is hard to know who is accountable.
Questions remain about how to reverse bad actions and where the final responsibility lies. Solving these legal and trust issues will decide whether AI agents become useful tools or create larger risks in crypto and finance.
