英国情報コミッショナー局（ICO）は、英公共機関におけるAI生成コンテンツが情報公開法の対象となる旨のガイダンスを発行した。これには、職員が使用したプロンプトや、それによって生成された出力内容も含まれる。
新たな規則により、政府機関はChatGPTなどのツールをどのように利用しているかという詳細な問い合わせに対し、情報公開の検討が義務付けられることになる。ICOのガイダンスでは、公式業務のためにAIを通じて作成された情報は、情報公開法の適用範囲内であることが明示された。
AIによるレポート AIによって生成された画像 事実確認済み
A recent podcast episode raised concerns that the UK government’s growing use of AI tools in public services—and potentially in elements of legislative work—could increase security and sovereignty risks tied to overseas providers.
India's Ministry of Electronics and Information Technology proposed persistent labels for AI-generated content on social media in a notice issued on April 21. The move amends IT Rules to enhance oversight on user-generated news. Feedback is invited until May 7.
AIによるレポート
The Japanese government announced on Friday it will establish a council of experts to discuss whether unauthorized use of sound data in AI-generated content emulating voice actors violates the Civil Code, amid advances in generative AI. The Justice Ministry panel will also address use of actors' images and present guidelines by July, as no legal precedent exists.
GCHQ has introduced a new AI system designed to detect cyber threats targeting critical sectors in the UK.
AIによるレポート
A Utah congressman has proposed the first federal legislation aimed at restricting artificial intelligence in toys marketed to young children. The measure would prohibit the manufacture and sale of such products in the United States. It comes amid growing concerns over safety, privacy and developmental impacts.
A German court has ruled that Google is liable for incorrect information in its AI Overviews, issuing a temporary injunction against the company. The decision from the Regional Court of Munich marks the first time an AI firm has been held accountable for its tool's outputs in this way. The ruling stems from a case brought by two Munich-based publishers.
AIによるレポート
Tim Berners-Lee presented his AI assistant Charlie during an event at SXSW London on Wednesday. The tool is designed to give users control over their personal data when interacting with large language models.