Autonomous Agent
The short answer · 19 words
In short
Software or hardware system that independently selects and executes actions within defined goals and limits without continuous direct control.
Flow-alerta.cab (2026): Autonomous Agent. https://flow-alerta.cab/en/glossar/autonomous-agent. Editorially reviewed on: 09 October 2026.
How much detail would you like?
Practical
Helps to delineate real scope for action from mere chatbots or fixed workflows. For companies, permissions, approvals and cancellation conditions are central.
Example scenario: An agent checks new support cases, retrieves permitted sources of knowledge and independently creates solutions, but may not change customer data.
Gradually awarded autonomy: first read tools and limited actions, then expand only after tests and traceable approvals.
At a glance
- Core
- For agents, autonomy means that the system has a certain room for manoeuvre and does not have to be controlled immediately for every single step. Classical agent research also connects agents with reactivity, goal orientation and interaction with their environment. Modern LLM agents are a possible form, not the definition of the term.
- Use
- Helps to delineate real scope for action from mere chatbots or fixed workflows. For companies, permissions, approvals and cancellation conditions are central.
- Example
- Example scenario: An agent checks new support cases, retrieves permitted sources of knowledge and independently creates solutions, but may not change customer data.
- Limits
- Confuse autonomy with unlimited freedom of action; automatically call chatbot agents; overly broad permissions; lack of logging and escalation paths.
In context
AI Agents
AI agents handle tasks over several steps and select tools and next actions within specified limits.
Tool Use / Function Calling
Function Calling allows a model to suggest tool calls that an application executes in a controlled manner.
Human-in-the-Loop
Human-in-the-loop means that people decide or examine in important places.
Questions about this term
Is every AI agent autonomous?
Not to the same extent.
Is an Autonomous Agent the Same as an LLM?
No.
Should an agent be given unlimited rights?
No, rights should fit the specific task.
All details and background
For agents, autonomy means that the system has a certain room for manoeuvre and does not have to be controlled immediately for every single step. Classical agent research also connects agents with reactivity, goal orientation and interaction with their environment. Modern LLM agents are a possible form, not the definition of the term.
ELI5 – Explained simply
You set goals, tools and limits. Within this framework, the agent decides for itself which next steps it will take.
Why it matters in practice
Helps to delineate real scope for action from mere chatbots or fixed workflows. For companies, permissions, approvals and cancellation conditions are central.
Example
Example scenario: An agent checks new support cases, retrieves permitted sources of knowledge and independently creates solutions, but may not change customer data.
Pitfalls and limits
- Confuse autonomy with unlimited freedom of action; automatically call chatbot agents; overly broad permissions; lack of logging and escalation paths.
Origin & history
Agent research of the 1990s described autonomy, reactivity, proactivity and social ability as central properties of intelligent agents. Modern agent AI revisits parts of these concepts with language models and tool access.
Use in mid-sized companies
- Gradually awarded autonomy: first read tools and limited actions, then expand only after tests and traceable approvals.
Related terms
- AI agents
- Tool Use
- Human-in-the-loop
- Multi-agent system
Matching Flow-alerta service
AI assistants & automationAI assistants & automation: test a suitable process with clear rights and human test points.Matching tool / use case
View use casesUse Case: Agent for support preparation with read access and human sharing.Sources
- www.cs.ox.ac.ukWooldridge & Jennings, Intelligent Agents: Theory and Practice (1995)
- doi.orgCambridge Core
Quellenprüfung: 09.10.2026. Definition und dokumentierte Einordnung abgeglichen; Beispiele beschreiben mögliche Anwendungen.
Editorially reviewed on: