Empfohlene Ressourcen
EU AI Act - Official Full Text
The EU Artificial Intelligence Act is the world's first comprehensive legal framework for AI. It establishes a risk-based approach to AI regulation, categorizing AI systems into prohibited, high-risk, limited-risk, and minimal-risk categories. The regulation sets requirements for high-risk AI systems including risk management, data governance, transparency, human oversight, and accuracy. It applies to providers and deployers of AI systems in the EU market.
OECD AI Principles (2019): the five principles adopted by 46+ countries and the G20
The 2019 OECD AI Principles are the first intergovernmental standard on AI, adopted by 46+ countries and endorsed by the G20. Five principles: inclusive growth, human-centered values, transparency, robustness, and accountability. This guide explains each principle and how they map to the EU AI Act and ISO 42001.
Kürzlich hinzugefügt
California SB 243: Companion AI Guardrails Act
California Senate Bill 243, signed into law on October 13, 2025 by Governor Gavin Newsom, makes California the first state to mandate specific safety safeguards for AI companion chatbots. The law takes effect January 1, 2026, and requires chatbot operators to implement critical safety measures around interactions with AI, particularly for minors. Key requirements include disclosure that users are interacting with AI, content guardrails preventing sexually explicit material for minors, suicide prevention protocols with crisis resources, and annual reporting to California's Office of Suicide Prevention. The law creates a private right of action for injured individuals.
Practices for governing agentic AI systems: OpenAI's seven safety principles
OpenAI's seven practices for safe agentic AI: accountability assignment, action ledgers, human approval gates, capability boundaries, staged deployment, reversibility design, and shutdown controls.
OWASP AI Bill of Materials (AIBOM)
OWASP's AI Bill of Materials (AIBOM) project establishes a standard format for documenting AI system components, training data sources, model provenance, and security configurations. Similar to how SBOMs transformed software supply chain transparency, AIBOMs aim to bring clarity to AI system composition, enabling organizations to track data lineage, model dependencies, and security configurations throughout the AI lifecycle.
C2PA Content Credentials Specification
The C2PA (Coalition for Content Provenance and Authenticity) Content Credentials specification establishes a technical standard for cryptographically binding provenance information to digital content. Led by Adobe, Microsoft, Intel, BBC, Truepic, Sony, OpenAI, Google, Meta, and Amazon, this standard enables verification of content origin, modifications, and AI generation across the media ecosystem.
AIGP certification: exam details, study guide, and what the IAPP credential covers
The IAPP AIGP (AI Governance Professional) certification covers four domains: governance foundations, risk management, technical controls, and organizational integration. 100-question exam, 2.75 hours, scaled scoring with 300 pass threshold. Aligned with EU AI Act, NIST AI RMF, and ISO 42001.