Would You Trust AI With 100% of Your Crypto? Two Experts Give Opposite Answers

100% versus 5%. Same question, polar opposite answers.

"What percentage of your portfolio could AI manage better today?" At NEARCON 2026, this question split two crypto heavyweights down the middle. Kraken co-CEO Arjun Sethi didn't hesitate: "One hundred." Dragonfly's Haseeb Qureshi was far more cautious: "Five percent."

Both men manage billions in crypto assets. So why did they land on completely different planets?

The 95% Problem: When 'Pretty Good' Isn't Good Enough

Qureshi's skepticism runs deeper than surface-level caution. "Something that works with money 90% of the time is unusable for actual economic activity," he argued. Even 95% reliability isn't sufficient.

His concern isn't theoretical. "It's a lot of nothing, nothing, nothing... then something, and then everything. And right now, we're still in the nothing phase."

The Dragonfly partner warned against the Twitter demo trap: "You want to be very cautious of trying to ingest your worldview of technology by reading Twitter hype people and watching Twitter demos." For major consumer platforms, he added bluntly, "You cannot do that sh**."

Qureshi sees a binary world: either AI systems are robust enough for meaningful capital, or they're expensive toys. There's no middle ground when real money is at stake.

The Speed Believer: 'Everything, in 6-12 Months'

Sethi operates from a fundamentally different premise. "The speed and the level of innovation... is exponential," he said. Where Qureshi sees reliability thresholds, Sethi sees rapid iteration closing gaps.

Kraken isn't just talking theory. The exchange is "already building agent-like capabilities for customers weeks and months away — not years away." This isn't a research project; it's active product development.

When pressed on whether he'd put all his crypto into an autonomous agent within a year, Sethi doubled down: "Everything. In the next six to twelve months."

That's not bravado from a startup founder. This is the co-CEO of an exchange processing billions in daily volume saying he'd trust AI with his personal wealth.

The Trust Spectrum: Where Do You Land?

The debate reveals crypto's broader philosophical split. Qureshi represents the "prove it first" camp — demanding near-perfect reliability before deployment. Sethi embodies the "iterate fast" philosophy — betting that defensive capabilities will scale alongside risks.

Both approaches have merit and massive implications. If Qureshi is right, autonomous finance remains years away from mainstream adoption. If Sethi is correct, we're months away from AI managing trillions in assets.

The Consumer Reality Check

For everyday crypto holders, this isn't academic. Major exchanges are already testing AI-powered features. Coinbase has portfolio rebalancing tools. Binance offers grid trading bots. The question isn't whether AI will manage crypto — it's how much control we'll cede.

Consider your own risk tolerance. Would you let AI:

• Rebalance your portfolio automatically?
• Execute stop-losses without confirmation?
• Make new investment decisions based on market patterns?
• Manage your entire crypto allocation?

Each step represents a different level of trust — and potential reward or catastrophe.