The EU passes the world’s first major AI regulation, creating a framework for Big Tech

New rulebook… The EU approved the first-ever comprehensive AI law to place guardrails around the explosive tech. The landmark law is set to go into effect in May. The EU’s effort to pass AI rules started three years ago (pre-ChatGPT era). The legislation separates tech into risk categories, ranging from “unacceptable” (aka: fully banned) to low risk. Companies that break the rules, including US tech giants, can be fined up to 7% of their annual revenue.

• Immediately no: The act bans AI applications that impinge on citizens’ rights. Think: facial-recognition databases trained on photos from the internet or CCTV, and emotion-recognition software in the workplace.

• Copywrong: AI-generated content will have to be labeled as such to prevent misinfo from deep fakes, and AI models must comply with European copyright laws. Companies like Microsoft and OpenAI are dealing with huge copyright-infringement suits. 

Pandora’s box… The AI Act comes as governments try to reel in the fast-expanding technology. Over a third of EU companies have adopted some form of AI, and 27% of Americans said they interacted with AI several times a day last year (picture: asking Dall-E to cut your ex from a pic). Leading players like Microsoft, OpenAI, and Meta have lobbied to help shape the rules. Past EU tech regulations have had significant effects on US tech companies:

• The EU’s Digital Markets Act (DMA) gives the union sweeping powers to govern anticompetitive tech practices. This month, the EU slapped Apple with a $2B fine, saying the biz suppressed music-streaming competition.

• Europe’s GDPR dictates how Big Tech cos handle user data (like: cookie-consent pop-ups). California introduced a similar privacy act shortly after GDPR passed.