The New AI Disclosure Rules Aren’t Just a Legal Issue — They’re an L&D Challenge
AI has quietly woven itself into everyday workflows. Copywriters use it to draft content, designers use it to generate images, HR teams use it to build training materials, and customer service teams rely on chatbots to handle routine questions. Most organizations treat this as a productivity shift. But as of this summer, it’s also a regulatory shift , one that touches far more employees than most leaders realize.
On August 2, 2026, two major transparency laws became enforceable: California’s AI Transparency Act (SB 942) and Article 50 of the EU AI Act. They were written by regulators, but the practical responsibility doesn’t fall on regulators. It falls on the people inside your organization who use AI every day. That’s where Learning & Development becomes essential.
What Actually Changed
California (SB 942)
California’s new rules focus on the companies that build generative AI. Providers with more than one million monthly users in the state must now offer free detection tools, give users the option to add visible AI disclosures, and embed hidden watermarks in AI‑generated images, video, and audio.
For most organizations, this isn’t a direct compliance burden. The responsibility sits with the major AI vendors, not the businesses using their tools. But it does mean your employees will start seeing new disclosure options and watermarking features inside the tools they already use and they need to understand what those mean.
The EU (Article 50)
Article 50 reaches much further. It doesn’t care how big your company is. It cares about what you’re doing.
If you build an AI system, you’re a provider with significant technical obligations. If you use AI such as a chatbot on your website, an AI writing assistant, an AI video generator, you’re a deployer, and you still have a clear duty: tell people when they’re interacting with AI or viewing AI‑generated content.
And that disclosure can’t be buried in a footer or hidden in a terms‑of‑service page. It must be clear, visible, and provided before the person engages with the content or tool.
The practical takeaway: Company size doesn’t exempt you in the EU, and “we didn’t build the AI, we just use it” doesn’t exempt you either.
Why This Is an L&D Challenge, Not Just a Legal One
Legal teams can write policies. They can’t stand next to every employee who opens an AI tool to draft a client email, build a landing page, or generate a training video. The gap isn’t policy; it’s knowledge and behavior. And closing knowledge and behavior gaps is exactly what L&D exists to do.
This shows up in everyday roles:
Marketing and content teams need to know when a disclosure is required and what “clear and distinguishable” actually looks like.
Customer‑facing teams need to understand disclosure obligations before a customer ever asks, “Am I talking to a bot?”
HR, consultants, ops, and MarCom need a baseline understanding of when AI involvement in a document or report should be flagged, especially for EU‑facing content.
Leadership and legal‑adjacent staff need enough fluency to recognize when a use case crosses into regulatory exposure.
None of this requires turning your staff into compliance officers. It requires basic AI literacy: what these rules cover, what they don’t, and where the judgment calls live.
Building AI Literacy Into Your L&D Strategy
A few starting points for any organization touched by these rules:
Map your AI touchpoints Before you can train people, you need to know where AI is being used. That includes documents, customer interactions, internal tools, and content creation. Most companies are surprised by how much they find.
Separate provider risk from deployer risk Most organizations fall into the lighter‑touch deployer category. Training should reflect that distinction, so employees understand what they’re responsible for and what they’re not.
Build disclosure into workflows, not just policies A policy nobody reads doesn’t change behavior. A prompt in a content‑review checklist or a standard line in an AI‑assisted chat tool does.
Make it role‑specific A website content editor needs different guidance than someone drafting client‑facing reports. Generic “AI 101” training rarely sticks. Targeted, use‑case‑based training does.
Revisit this regularly Guidance is still evolving, grace periods are ending, and more jurisdictions are likely to follow California and the EU’s lead. Treat this as an ongoing literacy effort, not a one‑time module.
The Bigger Picture
At their core, these laws ask a simple question: Does the person on the other end know they’re dealing with AI?
They reflect a broader shift toward transparency in digital work, a shift that benefits organizations, customers, and employees alike. Building that awareness into your L&D programming now ensures your teams are aligned with emerging norms and ready for the next wave of AI‑related requirements.
This post is intended as general awareness content, not legal advice. If your organization needs to assess specific compliance obligations under SB 942, the EU AI Act, or related laws, consult with legal counsel familiar with your use cases and jurisdiction.