Webot, the licensed exchange formerly known as Pionex.US, this week laid out a three-stage plan to bring AI into its built-in trading bots. The roadmap starts with AI that explains the product in plain language, moves to AI that watches a running bot and offers suggestions, and ends with AI that can act — but only within set boundaries. CEO Jay Hua said today's bots are 'blind' to shifting market conditions, which is why the exchange is pushing AI to help them read live data against existing settings.
From blind bots to AI eyes
Hua, a former vice president at Goldman Sachs, described the current generation of trading bots as reactive rather than perceptive. They follow rules, but they don't see when the market changes. The first stage of Webot's plan is an AI layer that explains how the product works and what each feature does. The second stage watches a bot in real time and suggests adjustments. The third stage lets the AI act on its own, but only inside boundaries the user sets.
That progression matters because it gives users a chance to trust the system before handing over control. Webot is licensed in 48 U.S. states and the U.S. Virgin Islands, and holds a MiCA Trading Platform Authorization in Ireland. The rebrand from Pionex.US to Webot followed the EU license, a sign the company is thinking beyond the U.S. market.
The safety gate
Alevtina Labyuk, chief strategic partnerships officer at BeInCrypto, warned that AI can gather context — news, sentiment, positioning — but it can't guarantee results. She said traders shouldn't hand over decision-making entirely. After recent exchange failures, she argued, the first gate is whether the venue is regulated and safe, especially in Europe, before letting AI move from fixed rules to advice to more autonomous action.
That's a pointed reminder for a sector that has seen its share of collapses. Webot's licensing status is a selling point, but Labyuk's point is that even the smartest AI doesn't replace the need for a trustworthy platform underneath.
Backtesting wish list
Modern Mining, a YouTube founder who has run Webot grid bots for about three years, wants AI backtesting after the bot is live. He'd like to see what the last 30 days paid and how tweaking the number of grids would have affected results. That's a practical ask — not a flashy feature, but one that helps users understand whether their bot is actually working.
Abay Beyshenkulov, partner at VComms, said venture money in crypto has been thinner than in other industries, but some is now looking at the overlap between Web3 and AI. Institutional capital is coming back, he said, even if it's selective. Labyuk added that Bitcoin has become less about crypto-native headlines and more about ETF flows, rates, and institutional money.
Modern Mining said Bitcoin had been stuck for months and then finally moved higher, but he didn't rule out a short-term pullback. The chat at the event, hosted by Webot, focused on the rebrand, risk tools, and how a bot should behave when volatility picks up. Those are the questions that will shape whether the AI roadmap actually gets used — or just gets talked about.
