生成AIパスポート syllabus revamped for Feb 2027 exam

The revision comes as generative AI spreads through both individual and organizational use, and as new risks surface — inappropriate use of generated content, over-reliance on AI, and information leaks through shadow AI, or unsanctioned AI tools used inside a company. GUGA frames the answer as human-centered: users should build ethics around how they work with AI, evaluate and verify what it produces, and take final responsibility for decisions.

The syllabus changes follow that logic. A section on AI agents has been expanded and the chapter structure reorganized, with contextual engineering and prompt injection added next to updated material on RAG (a method that lets AI pull in external documents) and MCP (a protocol for connecting AI to tools). What was previously taught as "information literacy, basic principles and AI social principles" is renamed "generative AI risk prevention, AI ethics and governance", with new items on typical risks, the mindset of user judgment and responsibility, and how to build organizational AI governance. Coverage also grows from basic uses to practical methods by job function — marketing, creative production, data analysis, app development — and by industry, through to evaluating and verifying AI output.

Whether the overhaul lands as intended may hinge on whether employers and certified staff actually follow the new material: the renewal test gives existing holders a path to show their literacy is current, but its uptake is the open question. For training and HR teams, the practical date to note is the February 2027 exam.