San Francisco — At NEARCON 2026, a sharp onstage debate captured a growing fault line in crypto: how soon autonomous AI agents should be trusted to move real money.
Kraken co-CEO Arjun Sethi and Dragonfly partner Haseeb Qureshi agreed on one point — that AI-driven agents will eventually manage capital — but they parted ways on timing and tolerance for risk.
Qureshi urged caution. “Something that works with money 90% of the time is unusable for actual economic activity,” he said, arguing even 95% reliability is insufficient. “It’s a lot of nothing, nothing, nothing… then something, and then everything. And right now, we’re still in the nothing phase.” He warned against taking viral demos at face value and urged skepticism of social-media hype, pointing to real-world failures of autonomous systems. For major consumer platforms, he added bluntly, “You cannot do that sh**.”
Sethi painted a much more optimistic picture. He described rapid, exponential improvements in AI and said Kraken is already building agent-like features for customers that are “weeks and months away — not years away.” He argued that as attack surfaces grow, so too will defensive capabilities, narrowing the gap between risk and utility.
The contrast came into sharp focus in a rapid-fire exchange. Asked what share of his own portfolio an AI could manage better right now, Qureshi said “five percent.” Sethi replied, “one hundred.” When pushed on whether he would entrust all his crypto to an autonomous agent within a year, Sethi didn’t hesitate: “Everything. In the next six to twelve months.”
Their clash highlights a larger debate in crypto between proponents who see autonomous finance as an imminent infrastructure shift and skeptics who view it as still experimental and fragile. The outcome will shape product roadmaps, regulatory stances, and how much capital people are willing to place in AI-run systems.
Further reading: Dragonfly recently raised $650 million amid a challenging market environment.
Read more AI-generated news on: undefined/news
