Agentic AI refers to artificial intelligence systems that are designed to exhibit agency—meaning they can make decisions, take actions, and pursue goals autonomously without direct human input. These AI systems are capable of operating independently within a set environment or task, making them more proactive than reactive. Unlike basic AI, which follows strict programming or responds to specific commands, agentic AI can assess situations, reason through options, and take steps toward fulfilling its objectives.

The term "agentic" is borrowed from psychology, where "agency" refers to an individual's ability to make choices and act independently. In AI, this means that the system has some level of autonomy, although the extent of autonomy can vary. Some agentic AI systems might work within narrow constraints (e.g., a self-driving car making real-time decisions within the rules of the road), while others could be designed to function in broader, more unpredictable environments.

While agentic AI can be powerful, it also raises concerns about control, safety, and ethics. Ensuring that these systems align with human values and don’t act in unintended ways is an important focus in AI research.