The EU and California Enforce Their AI Transparency Laws on the Same Day

My take: Today, August 2, 2026, two AI transparency laws take effect simultaneously: Article 50 of the EU AI Act and California's AI Transparency Act (SB 942). It is the first time two regulatory regimes of this scope have coordinated the same start date, and together they cover most of the markets where AI platforms operate.

The obligations are concrete. In Europe: chatbots must disclose they are machines at the start of each interaction, and AI-generated or AI-altered content (images, video, audio, informational text) must be marked in a detectable way. The maximum fine is 15 million euros or 3% of annual global revenue, whichever is higher. Systems already on the market before today have until December to comply with the technical marking requirements. In California: it applies to any generative AI provider with more than one million monthly users in the state, requiring C2PA-compatible provenance metadata in AI-generated images, video, and audio, plus a free detection tool.

For creators and businesses using AI to generate content: this is today, not the future. The direct question is whether you already have a clear process for disclosing or labeling the AI content you produce, or whether you are handling it without a defined policy.