⚖️Ethics & Regulation
Responsibility, rights and the rules around AI · 48 terms
AI Ethics Basics
5What it means for AI to be fair and to treat people well.
Making and using AI so it is fair, safe and honest.
When an AI is not fair to some groups of people, time after time.
Checking that an AI treats each group the same, and fixing it if not.
Your right to say who gets facts about you, and what they do with them.
A named person must answer for what an AI system does.
AI and Society
4How AI changes work, power use, privacy in public and war, far beyond any one app.
When AI takes over work that people were paid to do, task by task.
The power, water and stuff it takes to train and run AI.
Using AI to watch and track people. It can watch huge crowds.
Weapons that pick and attack a target. No human makes the call.
AI Transparency
5Being open about how an AI decides, what it is, and when it is being used.
Ways to show people why an AI made a choice.
A short page that says what an AI model is for, and where it falls short.
Telling people when they are talking to an AI. Or when AI made what they see.
A tool that guesses if writing was made by AI. It gives a score, not proof.
A signed record on a file that says who made it and how it was changed.
AI Misuse
4The harm that comes from fake, false or junk content made with AI.
A fake video, photo or voice of a real person. It shows them doing what they never did.
False or misleading content made or spread with the help of AI.
Junk content made with AI in bulk, mostly to get clicks.
A scam call that uses a cloned voice to sound like someone you trust.
AI Law
27The laws and rules that say what AI may do, and who is responsible.
The laws a government makes for how AI is built and used.
A 2025 package of changes to EU digital laws. It delayed parts of the AI Act.
A computer makes a choice about a person. No human checks it.
How a company picks its AI tools, checks them, and names who is in charge.
A firm's list of each AI tool it builds or uses. Each one has an owner.
EU AI Act
18The parts of the EU AI Act: who it names, what it bans or checks, and who enforces it.
The European Union's AI law. It sorts AI by risk and sets rules for each level.
How strictly the Act treats an AI use, from banned to free, and the tiers for general-purpose models.
The uses of AI the EU AI Act bans outright. Social scoring is one.
An AI used where one mistake could badly hurt a person.
AI that may be used freely, as long as people are told it is AI.
AI with little or no risk, like a spam filter. The EU AI Act adds no rules.
The EU AI Act's name for a model that does many tasks. Chat models are one.
The EU AI Act's top tier for the most powerful AI models. They carry extra duties.
Who the Act puts duties on: the firm that makes an AI system and the firm that uses it.
The maker of an AI system or model under the EU AI Act. It carries the most duties.
A firm that uses an AI system at work, under the EU AI Act. It did not build it.
What firms must do under the Act, from training staff to checks before launch and reports after.
Knowing enough about AI to use it well. And to see when it is wrong.
The check a high-risk AI must pass before it is sold or used in the EU.
Keeping watch on risky AI after launch. It catches faults in real use.
The duty to tell the state fast when a high-risk AI causes grave harm.
Steps AI model makers may sign up to. They help show the EU AI Act is met.
The bodies that check AI under the Act, and the safe spaces where firms can test with their help.
The EU body that oversees the AI Act. It keeps watch on the makers of big AI models.
The national body that checks AI on sale in one EU country. It can pull it.
An outside firm the EU trusts to check some risky AI. It works before sale.
A safe space where firms test new AI with a watchdog's help, before launch.
Copyright
4Who may train AI on owned work, how owners say no, and who owns what AI makes.
May AI makers train on books, art and music that others own?
A computer copies lots of text to find patterns. EU law allows it, with limits.
A way to say your work or chats must not be used to train AI.
Who owns what an AI makes. In the US, only the human part can be owned.
AI Standards
3Shared guides and standards for running AI with care, and the checks done before launch.
A free US guide. It helps you find and handle the risks of an AI system.
A world standard for running AI with care in a company. It can be certified.
A written check of how an AI could hurt people. It is done before launch.