New transparency obligations under the European Union's AI Act take effect, requiring providers and deployers of AI systems to clearly label AI-generated and manipulated content to mitigate risks of misinformation and fraud.