AI Agents to Revolutionize Blockchain Efficiency, Experts Say

Artificial intelligence is increasingly becoming a vital component of business operations, and blockchain technology is no exception. AI agents, which can automate various onchain tasks, are particularly beneficial for blockchain. However, the integration of AI agents into blockchain systems requires improvement to create more efficient use cases.

Recently, Cointelegraph Accelerator hosted a discussion with venture capitalists to explore the intersection of AI agents and cryptocurrency. The panel included Zoie Zhang, co-founder of Stealth Project, Fiona Ma, investment and research leader at DWF Ventures, and Samiz bayan, an investor at Draper Dragon. The conversation focused on how AI agents can enhance blockchain development and the potential game-changers in the industry.

Ma emphasized the need for more advanced AI agents that can make complex decisions and interact with multiple platforms. Currently, the market is dominated by basic and intermediate agents. She also noted that investors often view AI as a buzzword when founders fail to clearly explain its importance to their projects. Ma highlighted that AI agents can demonstrate specific use cases, such as monetizing user-generated content (UGC) through community-driven projects like Griffin AI or OpenAI Swarm.

Bayan added that decentralized finance (DeFi) and AI are a strong pair, not just for executing trades but also for monitoring positions and running automated tasks. Ma mentioned several DeFi projects that DWF Labs is involved with, including HeyAnon, which combines conversational AI with real-time data aggregation to manage DeFi operations, and AI16Z, which uses AI to redefine traditional fund management models by analyzing market sentiment and onchain data.

The speakers also discussed the growing attention to how AI agents work together, particularly through agentic workflows and coordination layers. These setups determine whether agents act in sequence or in parallel and how they share data and memory to achieve results. Zhang cited the example of Nethermind, an L2 run entirely by autonomous agents, where each agent registers itself onchain and transactions are governed by consensus among the agents.

In terms of institutional adoption, Bayan cited regulatory uncertainty and legacy systems as key barriers. He suggested a hybrid approach where institutions continue to rely on traditional systems while integrating blockchain in certain areas. He pointed to CARV as an example, which uses blockchain data for benefits and credentials but relies on offchain machine learning for computation.

Zhang noted that not every AI agent project needs to launch with a token right away. She stressed the importance of testing the market first, getting client feedback, and proving the use case before designing the tokenomics. Ma echoed this sentiment, emphasizing the need for AI projects built for long-term value with solid products, real revenue, and recurring cash flow. Bayan highlighted the importance of ease of use for both end-users and developers, suggesting that the next breakthrough moment will be when large Web2 companies start using blockchain-based compute effortlessly.

Zhang pointed to a trend of using AI agents to simplify workflows and integrate tightly with platforms that users already use, such as social media. She concluded that very soon, powerful products supported by AI-driven bots and streamlined AI agents will emerge.