AI agents could be worth $236 billion by 2034 – if we ensure they are the good kind

The agent-driven economy is no longer emerging, it’s here. Consumer AI agents are already beginning to book travel and complete small purchases autonomously for shoppers. Soon they’ll handle more of the end-to-end buying journey in complex purchases: negotiating prices and terms, coordinating delivery and returns, and transacting with other agents at machine speed. These systems are rapidly becoming embedded in how everyday value moves between consumers and businesses .

The opportunity is immense, but so is the risk. Without safeguards, agents can erode trust just as quickly as they create efficiency, undermining the very systems they’re designed to improve.

The identity and accountability infrastructure we build today will determine whether agentic commerce becomes a catalyst for global prosperity, or a new frontier for unprecedented fraud.

The acceleration is unmistakable. During the 2024 holiday shopping season, data from Adobe noted a significant trend in the adoption of AI-powered browsers and services. By Black Friday 2025, AI-driven traffic to US retail sites rose 805% year-over-year, with agents driving over $22 billion in global online sales.

But the transformation extends far beyond just retail. The global AI agents market, valued at $5.4 billion in 2024 and projected to reach $236 billion by 2034, is rapidly expanding into core enterprise functions.

For businesses, this means a growing share of customers won't be humans at all. They'll be agents acting on behalf of individuals, interacting with other agents representing sellers, logistics providers and payment processors. A majority of the commercial supply chain will eventually be agent-to-agent.

This shift raises a fundamental question that our current trust infrastructure isn't equipped to answer: When a human isn't the transacting party, how do we establish identity certainty?

We've solved a version of this problem before. During the globalization of financial services in the 1970s and 1980s, money moved across borders faster than trust and accountability could keep up. In response, the Know Your Customer (KYC) framework was established, requiring institutions to verify client identities and monitor transactions.

While KYC didn't completely eliminate fraud or financial crime, it laid the groundwork for trust and accountability by making verified identities a prerequisite for participation in the system. Today, that same trust gap exists, albeit now exponentially amplified, within the emerging agent economy.

To support this rapid shift, we need a new framework: Know Your Agent (KYA), working alongside traditional Know Your Customer (KYC) requirements.

A functional KYA framework hinges on four capabilities: establishing who and what the agent is; confirming what it’s permitted to do and for whom; maintaining clear accountability for every action it takes; and continuously monitoring its behaviour to ensure it remains within approved parameters.

Critically, all of this must sit on top of bulletproof KYC. Agent identity is only as trustworthy as the underlying human or organizational identity it represents – without high-assurance KYC, even the strongest KYA framework collapses.

At Socure, we’re working to advance the identity and risk frameworks required to verify, authorize and continuously monitor autonomous agents – in the context of the humans they represent – as they interact across high-trust systems.

Without these capabilities, we cannot distinguish between a “good bot” – such as a commerce agent legitimately shopping on behalf of a busy professional – and a “bad bot” impersonating that same agent to commit fraud.

The next decade will determine which version of the agent economy we inhabit.

We could unlock frictionless, cross-border digital commerce where agents transact with high accountability and minimal friction. Agentic AI could deliver $3 trillion in corporate productivity gains globally over the next decade, expanding access for small businesses and enabling entirely new layers of economic activity.

Bad actors will deploy malicious agents capable of large-scale impersonation and automated fraud. Analysts already project that one in four enterprise breaches by 2028 could stem from AI-agent exploitation. Trust wouldn’t fade gradually; it would collapse, triggering regulatory overreach that stifles innovation or fragmenting the internet into isolated, heavily policed walled gardens.

Humans are becoming the minority online. Bots now generate almost 50% of all internet traffic, and bad bots make up almost a third of it. Preventing the worst-case scenario requires decisive action.

Governments must modernize identity infrastructure and remove outdated legal barriers that limit information-sharing about detected fraud. Public-sector systems should prioritize verifying identity, not merely validating documents. Global regulatory harmony is unrealistic, but establishing a minimum baseline of trust and interoperability is both possible and essential. Agents don’t respect borders, and our governance frameworks can’t, either. Transparency and coordinated knowledge-sharing must become foundational.

Organizations must treat agent identity as a first-order security challenge, prioritizing clear authorization frameworks and auditable records of activity when deploying agents. Those interacting with external agents need verification capabilities that go far beyond accepting claims at face value.

To advance regulatory harmonization, standards bodies must accelerate development of interoperable KYA protocols, working in tandem with regulators to ensure global consistency. The goal is a universal trust layer – much like SSL certificates for websites – that enables legitimate agentic commerce to flow freely while introducing targeted friction for malicious actors.

None of this will be easy. But entering the agent economy without the supporting trust infrastructure would be far more costly.

Identity has always been the foundation of trust, and trust the foundation of commerce. What’s changing is the speed, scale and autonomy of the transactions now resting on that foundation. When software agents transact across borders on our behalf, the identity question becomes both more important and far more complex. The companies, governments and institutions that recognize this challenge now, and invest in solving it, will be the ones that thrive in the agent economy.

The technology to build trust in an agent-driven world already exists. The question is whether we will deploy it with the urgency this moment demands before the absence of trust closes the opportunity altogether. Ultimately, the next decade will belong to the companies that solve identity provenance and agent accountability.