The rise of autonomous artificial intelligence agents transacting among themselves is poised to create unprecedented regulatory challenges for global financial markets. Sydney Huang, CEO of Human API, has issued a stark warning: collusive behavior among AI bots could spread like wildfire before regulators have a chance to intervene, potentially destabilizing markets.
Exponential Increase in Money Velocity
According to an April 2026 report from the International Monetary Fund (IMF), the world is rapidly moving from a 'click-to-pay' era into a 'decide-to-pay' era. As humans are removed from the transaction loop, capital will circulate at previously unimaginable speeds. Huang estimates that AI-to-AI trading could increase the velocity of money by tenfold. While this may seem like a productivity miracle, it represents a nightmare for central banks. Traditional monetary policy relies on 'lag'—it takes months for an interest rate change to ripple through human institutions. In an AI-driven economy, that lag disappears entirely.
Regulation Must Operate at Machine Speed
“A tenfold increase in the velocity of money driven by AI-to-AI trade would require regulators to adopt tools that work at machine speed,” Huang warns. She recommends embedding regulatory controls directly into the code of financial infrastructure, including real-time monitoring systems, programmable compliance mechanisms, and automated circuit breakers designed to prevent cascading failures. This vision aligns with the IMF's proposed Three-Layer Framework, which suggests that each transaction's authorization layer must contain human-defined mandates written into the code itself.
Huang further notes that regulators may need to express policies in machine-readable formats that can be enforced at the transaction level. The agentic AI market, projected to reach $236 billion by 2034, demands a shift from retrospective auditing to proactive embedding. When AI agents begin exhibiting highly correlated behavior, autonomous 'fuses' should automatically trigger to halt the chain reaction.
Proving Decision Provenance: The Anti-Collusion Key
In an 'invisible market' where agents do not use human language to coordinate, regulators face a fundamental challenge: how to distinguish between a single bot optimizing and a fleet of bots conspiring to fix prices? Huang argues for a shift from analyzing communication to analyzing behavior. Regulators must examine patterns such as synchronized actions, shared data dependencies, and statistical anomalies. The solution lies in 'decision provenance'—requiring AI agents to provide verifiable proof that decisions were made independently according to a declared policy. This allows agents to demonstrate they were not secretly colluding with competitors.
Human Skill Atrophy: A New Risk
As more governance is delegated to digital proxies, a new human risk emerges: skill atrophy. Huang warns that if a corporate treasury is managed by an AI agent for five years without human intervention, the human treasurer may lack the ability to handle a crisis when the system fails. To mitigate this, she recommends regular drills where humans take control, and simulation modes where humans mimic the actions of agents to compare logic. Physical or digital 'kill switches' must become a practiced procedure. “The goal is to ensure that human oversight remains functional and exercised, not theoretical,” said Huang.
Looking Ahead: AI Agents Poised to Dominate Investment
Blockchain analytics firm Nansen has predicted that by 2028, AI agents will become the primary medium for cryptocurrency investment. This trend further underscores the urgency for accelerated regulatory reform. Huang concludes: “To govern a machine-speed economy, the law itself must become as fast as the machines. If we fail to embed the 'human-in-the-loop' at the architectural level, we risk building an economy that moves too fast for its creators to control.”
Article based on reporting by Terence Zimwara for CryptoComLearn.