AI regulation has moved from voluntary principles (Asilomar 2017, OECD 2019, UNESCO 2021) to binding law. The EU AI Act entered into force on 1 August 2024; it…
Researchers at Asilomar, California, agreed 23 voluntary principles on AI safety, transparency and human values.
Governments adopted the first intergovernmental AI standard: human-centred, transparent, robust and accountable AI.
A risk-based law: ban some uses, strict duties for high-risk AI, transparency for the rest.
Developers of the most powerful models had to share safety-test results with the government.
The world's first horizontal AI law took effect, with rules phased in over the next four years.
Social scoring, manipulative AI, facial-image scraping and emotion recognition at work and school.
Principle-based guidelines plus amended IT Rules: AI labels and 3-hour takedowns of unlawful deepfakes.
Agencies must harden systems against AI cyber threats; frontier-model review before release is voluntary.
The Digital Omnibus moved high-risk duties to December 2027 and August 2028, and added a ban on AI nudify tools.
Six AI firms signed a voluntary safety accord, and a Super Intelligence Force task force followed on 4 October.
Read the complete story, FAQs and key facts.