Venture Capital Eyes AI Agents in Crypto but Holds Back Investments

• Panelists at Consensus 2025 highlighted AI agents’ ability to automate trades and enhance crypto security through innovative incentive models.
• Executives agreed AI agents are not yet investable, focusing instead on infrastructure and long-term viability over immediate deployment.

Hong Kong’s Consensus 2025 conference hosted a panel discussion where crypto venture capital leaders expressed cautious optimism about artificial intelligence (AI) agents. Executives from Pantera Capital, Dragonfly, Maelstrom, and CMCC Global agreed the technology holds promise for crypto but remains too early-stage for direct investment.

Paul Veradittakit of Pantera Capital outlined practical uses for AI agents in crypto trading. He suggested these tools could automate complex trades across decentralized platforms, potentially boosting market activity.

“AI agents can process data rapidly and execute decisions without delays,” he said.

Similarly, Dragonfly’s “GM” proposed crypto-based incentives to improve AI safety, such as hackathons where developers compete for cryptocurrency rewards.

“This approach could align AI development with transparent, measurable goals,” he added.

Despite these possibilities, panelists revealed no active investments in AI agent projects. Akshat Vaidya of Maelstrom called the technology “not yet mature enough for capital deployment.” He speculated that future applications might include automating business operations, such as forming legal entities or managing compliance tasks.

We really enjoyed this session- 'Why are Crypto VC's Pivoting to AI' at the Emerging Tech stage of #Consensus2025 Day 2. 🎉

Different leaders such as Martin Baumann from #CMCC Global and Benjamin Schiller from #CoinDesk came together to discuss the immense growth of the #AI … pic.twitter.com/WRc1xztH4k

— OAX (@OAX_Foundation) February 20, 2025

“Imagine AI handling regulatory paperwork or even assisting in mergers,” Vaidya said. “That’s when the real shift happens.”

CMCC Global’s Martin Baumann confirmed his firm is monitoring AI agents but prioritizing infrastructure projects. GM echoed this stance, noting venture capitalists typically assess three- to five-year horizons before committing funds.

“Predicting how AI agents will evolve within crypto is challenging,” he said. Veradittakit acknowledged progress will require time, stating, “The groundwork is being laid, but adoption hinges on broader technological readiness.”

The discussion highlighted a consensus: AI agents could reshape crypto trading, security, and user engagement, yet practical implementation remains distant. Firms are weighing risks against uncertain timelines, opting to observe rather than act. For now, the fusion of AI and crypto stays theoretical—a vision awaiting tangible proof.