Fidelity research finds autonomous AI agents could transform financial markets, but warns that increased automation may not benefit all platforms or investors equally.
Fidelity Digital Assets' latest research highlights a significant shift in finance as artificial intelligence agents transition from supporting roles to functioning as autonomous market participants. These AI systems are expected to execute trades, arrange loans, manage portfolios, and process payments with minimal human intervention, operating continuously across multiple platforms.
The report underscores that increased automation does not guarantee greater value for every platform or institution. As AI agents optimize for cost and execution quality, they may rapidly switch between venues, concentrating benefits on platforms with robust liquidity, data access, and settlement infrastructure, rather than those with the highest transaction volumes.
By April 2026, AI agents had already conducted over 176 million blockchain transactions and settled approximately $73 million, illustrating the rapid expansion of automated financial activity. This measurable growth highlights the need for infrastructure capable of supporting high-frequency, low-value transactions between machines.
AI Agents as Independent Market Actors
Fidelity's analysis finds that AI agents will increasingly act as independent market participants, processing information and making decisions without human approval. Trading, lending, and portfolio management are identified as areas likely to experience the most immediate impact, given their reliance on rapid, data-driven decision-making and access to capital.
While payments generate high volumes, their thin margins mean they may not deliver lasting economic value to underlying infrastructure. The report cautions that competitive advantage will favor platforms offering seamless, reliable transactions for AI agents, rather than those simply pursuing higher activity.
Recent data from the Securities Industry and Financial Markets Association shows that electronic trading accounted for over 80% of U.S. equity market volume in 2025, reflecting the growing dominance of automated systems in daily operations.
Risks and Infrastructure Challenges
Fidelity's research warns that increased activity does not automatically translate into greater economic benefit for supporting networks. As AI agents transact in micro-increments, traditional banking infrastructure may struggle to handle the volume and granularity of machine-to-machine payments. Major providers such as Coinbase, Stripe, and Visa are investing in infrastructure to support these flows, which are largely invisible in standard trading metrics but could become a dominant source of settlement demand.
The report also notes a risk for public blockchains: some AI-driven financial activity may migrate to centralized networks operated by large technology and fintech firms, which offer better user experience, lower costs, and clearer regulatory frameworks compared to open networks.
As AI reduces the cost of replicating financial technology, the technology itself becomes less of a competitive advantage. Instead, network effects-liquidity, user trust, and regulatory clarity-become critical. Verification and auditability are essential for maintaining accountability and trust in autonomous systems.
Adapting Financial Infrastructure
The rise of autonomous AI agents will require banks, asset managers, and fintech firms to redesign infrastructure. Systems built for human users may not meet the needs of automated agents, which require digital identities, permissions, asset access, and auditable records. Early adopters could access new categories of financial activity, while laggards risk being bypassed by AI-driven flows.
Competition among platforms is expected to intensify as AI commoditizes underlying code. Value will concentrate in networks with deep liquidity, established user bases, and strong regulatory standing. Institutions that invest in infrastructure serving both human and machine clients are likely to be most successful.
Machine-to-machine payments represent a fundamental change in transaction processing. Unlike traditional payments, which are typically larger and less frequent, automated agent transactions occur in small increments at high frequency, placing new demands on settlement systems. The ability to verify, audit, and trust these automated flows will be a key differentiator as AI-driven finance evolves.
For context, the Securities Industry and Financial Markets Association reported that U.S. equity markets processed over 10 billion shares per day on average in 2025, with the majority handled by automated trading systems. This scale of automation highlights the importance of robust, transparent infrastructure as AI agents become more prevalent in financial markets.
