European Parliament Adopts AI Act Position: The World's First Comprehensive AI Law Advances

On June 14, 2023, the European Parliament voted to adopt its negotiating position on the Artificial Intelligence Act, sending the world's first comprehensive AI legislation into the final trilogue stage among the Commission, Parliament and Council. The Parliament version explicitly writes in requirements for generative AI: AI-generated content must be identifiable, and foundation-model providers must perform risk assessment, data governance and disclosure.

For the language industry, the "transparency" provisions carry the most direct meaning: machine translation and AI-generated text will face clear labeling requirements in the EU market. This means the "AI translation + human review" model must not only hold up on quality but also be "explainable" in process — which steps were done by AI, which by humans, all disclosable and traceable.

The deeper impact lies in the transmission of compliance culture: EU legislation has long exhibited the "Brussels effect" — multinationals, to unify global compliance, often apply the strictest rules across all operations. AI-transparency practices in language services will likely start with EU clients and gradually become a general expectation of enterprise clients worldwide.

For translation companies, now — before the rules bind — is the cheapest time to build internal AI-use policies and disclosure processes.