⏳ Two laws, one deadline: August 2, 2026

The August 2, 2026 AI Transparency Deadline

Two transparency laws take effect the same day. California's AI Transparency Act (SB 942) becomes operative for covered generative-AI providers, and the EU AI Act's Article 50 transparency obligations begin to apply. If your product uses generative AI, I can tell you exactly what has to change and handle the changes for you, on a fixed fee.

Prepared and handled personally by Sergei Tokmakov, Esq., California Bar #279869, licensed since 2011.

Fixed fee, no hourly billing Two business days from intake Attorney-reviewed, attorney-signed

Two ways I can help you meet the deadline

Start with the audit if you want to know where you stand. Move to the stack if you want the documents and disclosures actually built. Both are flat fees, confirmed in writing before any work begins.

Step 2 · Get it built

AI SaaS Launch & Compliance Stack

$2,500 flat fee
Best for: a company that wants the documents and disclosures drafted and implemented, not just a list of gaps.
  • Terms of Service, Privacy Policy, AI disclaimer and reliance language, and Acceptable Use Policy
  • Subscription and payment terms plus a data-processing framework
  • A product and vendor risk memo covering the transparency-disclosure obligations
  • Delivered in a private client workroom, with up to three revision rounds
Request the launch stack → Ask a question first by email
Not sure yet

Written Attorney Consultation

$240 flat fee
Best for: you want a written attorney read on whether either law even applies to you before you commit.
  • Send your question, a short summary, and a link to your product
  • Written attorney response on whether SB 942 or Article 50 is likely in scope
  • The main issues, risks, and practical next steps
Get a written consultation

The $2,500 stack is delivered through my "AI SaaS Launch Package" on PayPal; its document set is exactly the terms, privacy, disclosure, and acceptable-use work these transparency obligations require. Every engagement is subject to a conflict check, and I confirm the final fixed scope in writing before any work begins. Filing, registration, and ongoing monitoring are separate engagements.

These regimes are aimed at products that generate or manipulate content with AI, or that put users in front of an AI system. Read the list below:

  • Your product generates text, images, audio, or video using AI.
  • Users interact with an AI chatbot, assistant, or agent on your site or app.
  • You offer avatar, voice-cloning, face-swap, or deepfake-style features.
  • You use emotion recognition or biometric categorisation, or you publish AI-generated content on matters of public interest.
  • You serve users in the European Union (this pulls in EU AI Act Article 50).
  • You are a larger generative-AI provider (California's AI Transparency Act uses a user-count threshold to define a covered provider).
ℹ️If any of the first four describe you, plan on at least one of these regimes applying. Whether it does, and exactly how, depends on your specific product and facts. That is what the gap audit answers.
California · operative August 2, 2026

California AI Transparency Act (SB 942, as amended by AB 853)

The statute's own words are direct: "This chapter shall become operative on August 2, 2026." It targets larger generative-AI system providers, defined as covered providers by a user-count threshold. A covered provider must make a free, public AI-detection tool available, and apply latent ? machine-readable disclosures to content its system generates or alters, plus manifest ? visible disclosures on request.

Two later phases follow the first one: generative-AI hosting platforms and large online platforms on January 1, 2027, and capture-device manufacturers on January 1, 2028.

European Union · applies August 2, 2026

EU AI Act, Article 50 (transparency obligations)

Article 50 applies from August 2, 2026, and it is not limited to high-risk systems. Providers must inform people when they are interacting with an AI system, and must mark AI-generated or manipulated content in a machine-readable way so it can be detected. Deployers must disclose deepfakes, AI-generated text published to inform the public on matters of public interest, and emotion-recognition or biometric-categorisation systems.

⚠️One timing nuance worth knowing: under a May 2026 EU "AI Omnibus" provisional agreement, generative-AI systems already on the EU market before August 2, 2026 have until December 2, 2026 to meet the Article 50(2) machine-readable marking requirement specifically. The interaction-disclosure and deployer duties still track the August 2 date.
The exposure is real. Beyond civil penalties and enforcement, a fast-growing set of platform, app-store, payment-processor, and enterprise-customer requirements increasingly track these same transparency rules, so getting the disclosures right is quickly becoming a condition of doing business, not just a legal nicety.

Dates and statutory language above were checked against the California Legislature's bill text for AB 853 and the EU AI Act Article 50 materials on July 19, 2026. Legislative timelines can change; this is general information, not legal advice.

  • You request a package. Pick the gap audit or the stack above, or start with a written consultation if you are not sure the laws apply to you.
  • I run a conflict check and confirm scope in writing before any work begins. The fee is fixed, with no hourly billing.
  • You send a short intake: a link to your product, a description of the AI features, and where your users are.
  • I turn it around in two business days from your completed intake, with rush available if you are close to the deadline.
  • Everything is attorney-reviewed and attorney-signed. The AI Legal Analyst on this site can help you scope the matter, but the work product is mine.
Not sure which package fits, or whether either law applies to you? Open the AI Legal Analyst on this page, or email owner@terms.law.
Does this apply to my company if I only use a third-party AI, like an OpenAI or Anthropic model, inside my product?

Possibly. The transparency duties can attach to how you present AI features to your users, for example telling people they are interacting with an AI system or marking AI-generated output, even when the underlying model is someone else's. Where exactly the lines fall depends on your role and your specific product, which is what the gap audit sorts out. This is general information, not legal advice.

What is the difference between the California law and the EU law?

California's AI Transparency Act (SB 942) becomes operative for covered providers on August 2, 2026 and centers on a free public AI-detection tool plus latent and manifest disclosures on AI-generated content. The EU AI Act's Article 50 applies from the same date and centers on telling people they are interacting with AI, marking AI-generated or manipulated content, and deployer disclosures for deepfakes and emotion-recognition systems. Many companies touch both.

I already have a Privacy Policy and Terms. Do I still need this?

Existing terms and a privacy policy rarely address the specific AI-interaction disclosure, content-marking, and deepfake-notice duties these laws add. The gap audit tells you what your current documents already cover and what is missing; the stack rewrites and adds what is needed.

How fast can you turn this around before August 2?

Two business days from your completed intake for the gap audit, with rush available for a hard deadline. The stack is scoped in writing first and delivered in a private workroom with up to three revision rounds.

Can you guarantee my product will be compliant?

No, and be cautious of anyone who does. Compliance depends on the specific facts of your product and how these still-developing rules are enforced. What I provide is an attorney-prepared assessment and the documents to close the gaps I identify, confirmed in writing.

Disclaimer. This page is attorney advertising and general legal information about the California AI Transparency Act (SB 942) and the EU AI Act, not legal advice, and it does not create an attorney-client relationship. Statutory dates and requirements were verified against primary sources on July 19, 2026 and can change as regulations, guidance, and amendments develop. Whether either law applies to you, and what it requires, depends on the specific facts of your product. An attorney-client relationship is formed only after a conflict check and a written scope confirmation.

More on AI law from me