Industry executives at the Wyoming Blockchain Symposium 2026 have raised alarms over the potential for AI agents to be exploited in large-scale cyberattacks, while also debating the dangers of granting these systems excessive control over user data and financial assets.
Ryan Kirkley, CEO of Global Settlement Network, warned that AI agents could be weaponised to target financial systems on a massive scale, stating: “Now I can have an agent that goes after everyone.” He argued that current approaches to agentic AI assume agents are inherently trustworthy, a perspective he described as “a really fatal flaw.”
Bill Laboon of the Web3 Foundation highlighted how AI’s capabilities could lower barriers for malicious actors, noting: “This also means that there is less friction for the bad guys.” He added that even advanced language models occasionally produce false information, raising concerns about reliability.
Fahmi Syed of the Midnight Foundation stressed the need for strict parameters to govern AI agents, calling the idea of “giving all that power to a single agent” a “scary concept” without safeguards. Richard Shorten of Silvermine Capital Advisors echoed these concerns, saying the rapid evolution of agentic AI has outpaced public understanding.
The discussions come amid growing scrutiny of AI’s role in finance, with Kirkley questioning accountability in scenarios where agents act unlawfully: “If your agent goes and does something illegal… where is the ultimate decision maker?”
The Block, which hosted the panel, is majority-owned by Foresight Ventures, a firm with Bitget as its anchor limited partner. However, the symposium’s focus remained on technical and regulatory challenges rather than investment structures.
Written by Steven Cook
Regulation Desk
