The Competition and Markets Authority published research on 9 March 2026 examining how agentic AI (AI systems capable of taking actions autonomously on behalf of users, such as booking, purchasing, or negotiating) could affect consumers, how existing law applies to those interactions, and what businesses should do to manage the associated risks. The publication, titled 'Agentic AI and consumers', was accompanied by a separate piece of guidance on how businesses can benefit from agentic AI while still meeting their legal obligations. The CMA's research explores how agentic AI is being used today and what the future could hold as the technology becomes more capable. The authority's interest in the area reflects a broader concern that AI agents acting on behalf of consumers could create new information asymmetries, reduce consumer control over purchasing decisions, and introduce novel liability questions around who is responsible when an AI agent acts in a way that harms a consumer or a third party. The research sits alongside a separate body of consumer research finding that roughly one in three consumers have already used an AI agent for a customer service interaction, with adoption rising to approximately half among younger consumers. That data, drawn from a survey of US consumers published in October 2026, also found that consumers using AI tools are more likely to challenge company representatives on policy, escalate disputes, and introduce potentially inaccurate AI-generated information into interactions. The CMA's dual publication of consumer research and business guidance indicates that the authority views agentic AI as a near-term consumer protection issue requiring attention now, rather than a purely speculative future concern.
Why this matters
The CMA's decision to publish both research and accompanying business guidance on agentic AI signals that the authority views existing consumer protection and competition frameworks as applicable to AI agent behaviour without waiting for new primary legislation. This is significant because it means businesses deploying agentic AI in consumer-facing contexts face live compliance obligations today, not just future regulatory risk. The research also stakes out the CMA's analytical territory ahead of any future market study or enforcement action in this space, establishing the authority's framework for thinking about AI agent-driven harm. As agentic AI adoption accelerates, the question of who bears legal responsibility when an AI agent makes an erroneous purchasing decision or misrepresents a product creates genuinely novel liability questions that sit at the boundary of contract, consumer protection, and tort law.
On the Ground
Businesses deploying AI agents in consumer-facing settings will need legal advice on how existing consumer law obligations apply to automated purchasing or service interactions, including questions of informed consent, liability for agent error, and transparency requirements. Technology and commercial lawyers will be reviewing AI vendor contracts and data processing agreements to ensure appropriate allocation of liability where an AI agent causes consumer harm. Regulatory lawyers advising on CMA guidance will be assessing what compliance steps the business guidance requires. A trainee on an AI governance matter would assist with drafting AI governance policy documents, reviewing technology licence agreements for relevant indemnity and liability clauses, and preparing regulatory impact assessment memos summarising the CMA's published guidance against a client's current AI deployment.
Interview prep
Question you might get
“The CMA has published research and guidance on agentic AI and consumers. What legal questions does the widespread use of AI agents raise, and how should a business deploying an AI agent in a consumer-facing context approach its compliance obligations?”
Sign up free to see the full answer
A model answer you can lift into an interview — how to frame this story for a partner.
Sign up freeMy notes
saved