FCA Warns: Agentic AI and Tokenized Money Are Set to Disrupt the Financial System

The UK’s Financial Conduct Authority (FCA) has just issued an unprecedented warning: the convergence of agentic artificial intelligence and programmable money represents a systemic rupture for financial markets.

AI agents capable of executing autonomous transactions with tokenized assets — with no human intervention. This scenario, long confined to science fiction, is now at the heart of the UK regulator’s concerns.

Caught between historic opportunity and systemic risk, the FCA is calling for a fundamental overhaul of the regulatory framework. Here is what that means in practice for the crypto and financial ecosystem.

Agentic AI: When Machines Take Control of Financial Transactions

Agentic AI refers to artificial intelligence systems capable of making decisions and acting autonomously to achieve defined objectives — without requiring human validation at every step. Applied to finance, this technology opens the door to agents that can manage portfolios, execute orders, negotiate contracts, and interact with DeFi protocols in real time.

The FCA pinpoints precisely this tipping point: when these AI agents operate with programmable money or tokenized assets — stablecoins, CBDCs, tokenized RWAs — the speed and scale of transactions reach a level that current surveillance infrastructure simply cannot absorb. An AI agent could theoretically execute thousands of operations per second across tokenized markets, making any reactive human oversight virtually impossible.

The regulator also highlights the risk of financial decision-making becoming concentrated within a small number of AI models. If multiple institutions rely on similar architectures, a shared bias or common failure could trigger coordinated and amplified market movements — a systemic flash crash on an unprecedented scale.

Tokenized Money: The Fuel That Amplifies Agentic Risk

Tokenized assets — whether central bank digital currencies (CBDCs), stablecoins, or real-world assets (RWAs) on blockchain — represent the ideal substrate for agentic AI. Their programmable nature, near-instant settlement, and interoperability with smart contracts allow AI agents to operate within a fully automatable environment.

The FCA points to a fundamental paradox: the very characteristics that make these assets attractive — speed, programmability, the absence of intermediaries — become vectors of risk when coupled with autonomous agents. A misconfigured smart contract executed by an AI agent on a tokenized RWA market could generate massive losses before any human has the chance to intervene.

The UK regulator is therefore calling for a revision of liability rules: who is legally accountable for a decision made by an AI agent — the model’s developer, the financial institution deploying it, or the end user? This question currently has no clear answer in virtually any jurisdiction, including within the European Union despite the AI Act.

What the FCA Is Demanding: Regulation That Matches the Disruption

Faced with these risks, the FCA is not content to simply observe. The regulator is actively pushing for the development of new governance frameworks specifically designed for agentic AI, including requirements for the traceability of algorithmic decisions, automatic circuit-breaker mechanisms, and enhanced reporting obligations for institutions using these technologies.

The FCA also stresses the urgent need for international coordination. Agentic AI and tokenized assets know no borders: a fragmented regulatory approach across jurisdictions would immediately create opportunities for regulatory arbitrage, driving the most aggressive players toward the least restrictive environments.

For the crypto ecosystem, this regulatory signal cuts both ways. On one hand, it validates the trajectory of large-scale tokenization of financial assets as an unstoppable structural trend. On the other, it signals an incoming wave of regulatory constraints that will redefine market access conditions for DeFi protocols, stablecoin issuers, and financial AI agent developers — with direct implications for the competitiveness of established players relative to new entrants.