What is an AI agent?
An AI agent is an AI system that pursues goals by planning and executing multi-step actions with limited human intervention: calling APIs and tools, browsing, writing to business systems, or completing transactions. The step from chatbot to agent is the step from content risk to action risk — an agent’s mistake is not a wrong answer but a wrong booking, payment or contract communication.
The legal core in three questions
- Attribution: whose declaration is the agent’s action? Under Turkish contract law, generally the deploying company’s — authority limits belong in system design and counterparty terms;
- Oversight: Article 14-grade oversight must be designed for systems that act, not just suggest — approval gates for consequential actions, kill switches, action logs;
- Classification: an agent operating in an Annex III domain is a high-risk system; one interacting with consumers triggers Article 50 disclosure from August 2026.
Go deeper
Full analysis in our canonical entry AI agent (agentic AI) and the article Agentic AI: authority and liability.