AI & Tech One-page note UK only
AI Agents Explained
AI that doesn't just answer, it does things for you. Here's what AI agents are, what they can do, and how to use them safely. Save this page.
Good to know
Sources: International AI Safety Report (2026), UK NCSC (2025). Checked Nov 2026.
What's on the page
The note as text, so you can read it, copy it or use a screen reader.
Chatbot vs agent
- Chatbot: answers your questions
- Agent: takes actions to reach a goal
- You give the goal in plain words
What agents can do
- Use tools like a web browser
- Click, type and fill in forms
- Do several steps with little help
The limits today
- Still make basic mistakes
- Can make things up
- Struggle with long, complex tasks
The risks
- Acts on its own: harder to stop mid-mistake
- Hidden text on pages can trick it
- This "prompt injection" may never be fully fixed
Tips and mistakes
- Watch it work; check before it acts
- Give it only the access it needs
- Handle payments and passwords yourself
- Letting it run unsupervised on your accounts
- Trusting a tool that "blocks all attacks"
Sources
- International AI Safety Report 2026 (3 Feb 2026)
- NCSC: Prompt injection is not SQL injection (it may be worse) (8 Dec 2025)
- Anthropic Help Center: Getting started with Claude in Chrome
- Covington (Inside Global Tech): International AI Safety Report 2026 examines AI capabilities, risks and safeguards (10 Feb 2026)
How we checked this
Every fact on this note comes from the 4 sources listed here, checked against official or peer-reviewed sources where possible. AI tools help with drafting; a person checks every fact against the sources. How we research
Spotted a mistake? Tell us and we'll fix it and log it on the corrections page.