How countries regulate AI
Different places are choosing different ways to govern AI. The EU uses risk levels; others rely more on existing laws and targeted rules. The rules keep changing.
In 60 seconds
How countries regulate AI
Different places are choosing different ways to govern AI. The EU uses risk levels; others rely more on existing laws and targeted rules. The rules keep changing.
Why regulate AI at all?
- To protect people from harm, unfair treatment and loss of privacy.
- To make companies explain and take responsibility for high-impact systems.
- To build public trust, so useful AI can be used safely.
- To keep competition fair and support innovation.
The European Union: sorting by risk
| Risk level | Basic idea | Examples of the kind of use |
|---|---|---|
| Unacceptable risk | Banned | Certain manipulative systems and social scoring of the kind the law prohibits |
| High risk | Allowed with strict requirements such as risk management, data quality, documentation and human oversight | Some uses in hiring, education, credit and essential services |
| Limited risk | Transparency duties | Telling people they are talking to a chatbot; labelling certain AI-generated content |
| Minimal risk | Few or no new obligations | Spam filters, AI in video games |
Other approaches, in broad terms
| Place | Broad approach (high level) |
|---|---|
| India | No single, comprehensive AI law so far. The government relies on existing laws, such as the IT Act and its rules and the DPDP Act, along with advisories, guidelines and programmes to build AI capacity. Rules on topics such as labelling AI-generated content have been discussed. |
| United States | No single, comprehensive federal AI law. A mix of existing laws enforced by agencies, executive actions that have changed between administrations, and a growing number of state laws. |
| China | Specific rules for particular uses, such as recommendation algorithms, deepfakes (deep synthesis) and public generative AI services, including requirements such as labelling and security reviews. |
| International | Shared principles, such as those from the OECD and UNESCO, and global summits where countries discuss AI safety and cooperation. |
- “Country X has no rules, so anything goes”
- “One law will fix all AI problems”
- “Regulation always kills innovation”
- Existing laws on privacy, fraud, consumer protection and discrimination often still apply
- Rules, standards, company practices and public awareness all work together
- Good rules can build trust; bad rules can create confusion — design matters
Watch and read more
Lab
A risk classification and short policy comparison for three AI uses.
The problem
Choose three AI uses from daily life (for example: crop-disease detection, exam proctoring, a customer-service chatbot, a loan-approval tool, a face-recognition gate).
- For each, decide its risk level using the EU's four-level idea and explain why.
- List two protections you think each should have (for example human review, labelling, testing for bias).
- Using official or trusted sources, write two sentences each on how India and one other place broadly approach such uses. Note your sources and the date.
- Write a short paragraph: which approach do you find most sensible for your country, and what trade-off does it make?
You are done when
Hard questions
Try to answer before you reveal. If you can answer these, you understood the lesson.
Q1Why do many regulators focus on how AI is used rather than on the technology itself?Reveal
Because the same model can be harmless in one use and harmful in another. A text generator writing poems is low risk; the same model scoring job applicants can affect lives. Regulating uses lets rules target real harm without blocking every application of a general tool. The challenge is that general-purpose models can be used in so many ways, which is why some rules also address the models themselves.
Q2What problems can arise when different countries have very different AI rules?Reveal
Companies selling globally must follow many rulebooks, which raises costs, especially for small firms. Some may avoid certain markets or build separate versions. Gaps between countries can also let risky uses move to places with weaker rules. On the other hand, different approaches let countries learn from each other. International cooperation on shared principles and standards helps reduce the friction.
Questions people ask
Does the EU AI Act apply to Indian companies?
It can, if they place AI systems on the EU market or their systems' outputs are used in the EU, depending on the details. Companies in that situation should get qualified legal advice.
Is there an AI law in India?
India has not passed a single comprehensive AI law so far. Existing laws still apply to AI uses, including the Information Technology Act and its rules and the DPDP Act, and the government has issued advisories and guidelines. Check official government sources for the latest position.
Why don't all countries just use the same rules?
Countries have different legal systems, values, economic goals and views on the role of government. They also cooperate through international groups, and many share basic principles like safety, fairness and transparency.
Who decides whether an AI system is high risk?
Under the EU approach, the law itself lists categories of high-risk uses, and companies must assess their systems against them, with regulators overseeing. Other places use different methods.
Lesson test
5 questions. Get 3 right (60%) to pass and complete this lesson.