Fidelity crypto catch: AI may not enrich token holders
Fidelity Digital Assets warns that AI could flood crypto with apps and agent activity without delivering lasting value to token holders. Its fidelity crypto analysis says cheaper AI development may shift advantages to liquidity, trust and banks—not public chains—so more machine transactions may not mean higher token prices.
The crypto arm of Fidelity Investments framed the issue plainly in a Wednesday report by analyst Max Wadington: the question is less how much activity AI generates and more who captures the economic value. That warning lands as AI agents already settle meaningful onchain volume and major payments firms race to build machine-to-machine rails.
For readers tracking how tech hype meets hard risk, our True Crime & Unsolved Mysteries hub covers related investigations into exploits, fraud and unexplained digital losses.
Key Takeaways
- Fidelity Digital Assets says AI may boost crypto apps without guaranteeing demand or product-market fit.
- AI agents could prefer banks and fintechs over public blockchains for cost, performance and regulatory clarity.
- More transactions may enrich stablecoin issuers and service providers more than base-layer tokens.
- AI can also accelerate discovery of smart-contract and infrastructure vulnerabilities.
Why does Fidelity Digital Assets doubt AI crypto value?
AI is a rising crypto investment narrative because autonomous agents can buy data, computing power and transact without humans. A Keyrock report cited by CoinDesk found AI agents settled more than $73 million across roughly 176 million blockchain transactions in the year through April. Coinbase, Stripe and Visa are developing competing machine-to-machine payment systems.
Wadington argued that as AI lowers barriers to development, competitive edges may move to liquidity, distribution, security, trust and regulatory integration rather than technology alone. Alchemy CEO Nikil Viswanathan has said “crypto was built for AI agents, not humans,” but Fidelity stresses that cheaper builds do not equal durable demand.
Even when agents use blockchains, micropayments often produce low fees and can shift to Layer 2 networks or off-chain settlement. Coinbase’s x402 largely settles in Circle’s USDC, illustrating how stablecoin issuers and service providers may capture upside that base-layer tokens miss. Fidelity pointed to trading, lending and borrowing as stronger value-capture paths. Details are in CoinDesk’s report on the Fidelity Digital Assets note.
Could AI agents skip public blockchains entirely?
Fidelity said banks, fintechs and technology companies could offer lower costs, better performance, regulatory clarity and established distribution. A boom in AI-driven economic activity therefore does not guarantee a boom in blockchain activity.
Separately, infrastructure builders still argue public chains can host the agent economy. Crypto.news summarized The Graph’s view that onchain agents need identity, structured data and payments—via standards such as ERC-8004, Subgraphs with Model Context Protocol, and x402 micropayments in USDC—to run an autonomous “agent loop.” Forbes Technology Council also argued blockchain-based governance can make agent decisions auditable as enterprises deploy agents at scale.
What security risks does AI create for crypto protocols?
Fidelity warned AI can make vulnerabilities easier to find and exploit in smart contracts, key management, bridges and oracles. Anthropic’s Mythos model has pushed parts of the industry to rethink how AI accelerates vulnerability discovery. Defensively, Kraken parent Payward joined Anthropic’s Glasswing project for restricted access to Claude Mythos 5 to find and fix bugs before attackers do.
The firm’s catchline is stark: the risk is not that AI fails, but that it succeeds while crypto captures only a fraction of the value. For fidelity crypto watchers, activity metrics alone are no longer enough—where fees, trust and settlement land matters more.