AI Agent
An autonomous AI system that can perceive its environment, make decisions, and take actions to achieve goals — like managing your email, scheduling meetings, or monitoring data.
Why it matters
An AI agent goes beyond answering a question; it takes steps to get a goal done, like searching, using tools, and acting on results. It matters because this is where AI shifts from a smart chatbot to something that can actually complete multi-step tasks for you. It also raises the stakes: an agent that can act needs clear limits and a human watching what it does.
A concrete example
Instead of just suggesting a trip itinerary, an agent might search for flights, compare a few options, check a calendar, and draft a booking summary, chaining several actions on its own. That autonomy is powerful for busy workflows, but it's also why you'd want it to confirm before spending money or sending anything. Give agents room to work, but keep a hand on the wheel.
How to use it
Start with the agent proposing and you approving, and only widen its authority once you have watched it work. Define precisely which actions it may take alone, what it must escalate, and log everything it does — the point of a log is that you find out what happened before someone else tells you. Anything involving money, external communication or an irreversible change belongs behind an approval step for far longer than feels necessary.
The common mistake
Giving an agent broad autonomy because the demo worked. A drafting assistant that gets something wrong produces a draft you delete; an agent that gets something wrong has already sent the email or cancelled the booking. Autonomy multiplies both usefulness and consequence.
Related terms
Agentic AI
AI systems that operate autonomously over extended tasks — planning, executing, and self-correcting without step-by-step human guidance. Unlike chatbots, agentic AI sets sub-goals, uses tools, and adapts its strategy based on intermediate results.
Tool Use / Function Calling
The ability of AI models to invoke external tools, APIs, or functions during a conversation. Enables AI to perform real actions — searching the web, querying databases, sending emails, or running calculations — not just generate text.
AI Orchestration
Coordinating multiple AI models, tools, and data sources in a unified pipeline. An orchestration layer manages prompt routing, context passing, error handling, and output aggregation across different AI services.
AI Disclosure
Stating that AI was used in producing a piece of work, where a policy, client or publisher requires it. Distinct from permission: some contexts allow AI use but require it to be declared.
ATS (Applicant Tracking System)
Software used by employers to filter and rank job applications by scanning resumes for keywords, formatting, and relevance. AI resume builders optimize output to pass ATS screening.
Benchmark
A standardized test or dataset used to evaluate and compare AI model performance. Common benchmarks include MMLU (knowledge), HumanEval (coding), and MT-Bench (conversation). Helps users choose the right model for their needs.