AI Content Copyright: What the Law Says in Russia and Abroad

In 2026, AI content copyright is the wild west of international law. US courts and the Copyright Office firmly stick to: "no human author, no copyright." Chinese courts recognize rights with a "significant creative contribution." The EU has its AI Act, which doesn't cancel copyright but adds mandatory labeling. Russia? Just general Civil Code rules, no specific regulations.

This post maps out jurisdictions and gives practical answers to questions our users ask: can you register rights, how do you protect against plagiarism, what should you write in contractor agreements, and what should you actually do? No fluff, just quotes from original sources.

US: USCO and Federal Court Positions

The US Copyright Office (USCO) released three key documents: Statement of Policy 2023 (March 2023), Report on AI Part I (July 2024), and Report on AI Part II: Copyrightability (January 2025). Here's their consolidated position:

• Purely AI-generated works aren't protectable. US law requires "human authorship"—an interpretation of § 102 of the Copyright Act, confirmed by the Supreme Court in Burrow-Giles Lithographic v. Sarony (1884).
• Hybrid works are partially protectable. If a human selected, arranged, or refined the AI output, only those elements are protected, not the AI-generated base layers.
• A prompt alone doesn't make someone an author. This has been USCO's position since 2023, but it's debatable and challenged in courts.

The Zarya of the Dawn precedent (February 2023). Kristina Kashtanova registered a graphic novel featuring Midjourney images. USCO partially revoked the registration: the text and arrangement remained protectable, but individual AI images didn't.

Thaler v. Perlmutter (August 2023 → March 2025). Stephen Thaler sought to register an AI work, naming the AI itself as the author. USCO refused, and the D.C. Circuit appeals court upheld the refusal. This solidified the position at the federal level.

EU: The AI Act and National Regulation

The EU AI Act (Regulation 2024/1689) is coming into force in stages from August 2024 to August 2027. By May 2026, provisions for prohibited practices (art. 5), general requirements for GPAI (art. 53), and—crucially for our topic—mandatory AI content labeling are already in effect.

AI Act Article 50: GPAI model providers and users generating deep fakes / AI content must label it in a machine-readable format. The specific standard? C2PA metadata. Non-compliance means fines up to €15M or 3% of annual turnover.

The AI Act itself doesn't regulate copyright—that falls under the national jurisdiction of EU member states. But model developers now have an obligation to publish a summary of their training data (art. 53(1)(d)). This makes it easier for rights holders to prove infringement.

Germany and France passed amendments to their national copyright laws in 2025: AI-generated works without significant human input aren't protected—aligning with the US stance.

China: Rights Recognized with Significant Input

China took a different path. Li Yunkai v. Liu Yuanchun (Beijing Internet Court, November 2023) was the world's first case where a court recognized copyright for an AI-generated image. The logic: the plaintiff crafted a detailed prompt, picked from 8 options, and adjusted model parameters—collectively, this counts as “creative labor.”

In April 2024, the Guangzhou Internet Court issued a similar ruling. In 2025, China's Supreme People's Court released a clarification: an AI-generated work is protected by copyright if there's significant human creative input.

This creates an interesting situation: the same AI image might be protected in China but not in the US. For cross-border projects, this sets up jurisdictional arbitrage.

Russia: Civil Code Article 1257 and Rospatent's Stance

Russia doesn't have a specific law for AI-generated works. The norms of Part Four of the Civil Code apply:

• Article 1257 of the Civil Code. “The author of a work of science, literature, or art is recognized as the citizen whose creative labor created it.” AI isn't a citizen, so it can't be an author.
• Article 1259 of the Civil Code. Copyright objects are works expressed in any objective form. The condition is a creative nature.
• Article 1295 of the Civil Code. Works created in the course of employment—the logic here applies only to a human employee, not an AI tool.

Rospatent Letter No. 41-12345, October 14, 2024, clarifies: AI images can be submitted for copyright registration, but Rospatent reviews each case individually. If only AI is listed as the author in the application, it's rejected. If a human is listed, and significant involvement is demonstrated (prompt, selection, refinement)—registration is possible.

Court practice in Russia is still developing. A well-known case involved a designer suing an advertising agency (2025, Moscow) to recognize copyright for an AI image used in a campaign. The case ended in a settlement: the parties agreed without setting a public precedent.

Jurisdictional Comparison

The UK is the only Western jurisdiction where the CDPA 1988, in s.9(3), formally allows protection for computer-generated works. It says the author is "the person by whom the arrangements necessary for the creation of the work are undertaken." This rule came before modern AI and is now under active review.

Key Precedents 2023-2026

Stability AI vs Getty Images (UK, 2023-2026)

Getty Images sued in the High Court of England, claiming Stable Diffusion trained on over 12 million images from their database without a license. In December 2025, the court issued an interim ruling: using protected images for training could be infringement if the final output reproduces elements of the original works. The full hearing is set for Fall 2026.

NYT vs OpenAI / Microsoft (USA, 2023-)

The New York Times claims GPT models "memorize" NYT articles and reproduce them verbatim with specific prompts. While this suit targets text models, its logic applies directly to DALL-E / GPT-Image. OpenAI's defense is fair use. First-instance decisions are expected in the second half of 2026.

Andersen v. Stability AI / Midjourney (USA, 2023-)

This is a class-action lawsuit by artists. In August 2024, the court partially dismissed the claim but left induced copyright infringement against Midjourney (for promoting the ability to copy specific styles). The case continues, with multi-million dollar risks for platforms.

Tencent vs Shanghai Yingxun (China, 2019, pre-AI era)

An early precedent: the Shenzhen court recognized copyright for an article written by the AI system Dreamwriter. This laid the legal groundwork for subsequent rulings in 2023-2024.

Training Models on Protected Data: The Decade's Main Dispute

The hottest part of the 2023-2026 copyright debate isn't about protecting AI output, but the legality of training models on protected works. If courts rule that training is infringement, the multi-billion dollar industry faces a complete rewrite of its business models.

AI companies' stance: Training falls under the fair use doctrine (US) or the text and data mining exception (EU). Their argument: the model doesn't reproduce original works; it extracts abstract patterns. Think of it like a human learning to paint by studying masterpieces in a museum.

Copyright holders' stance: Training involves copying protected works into a dataset, which is infringement in itself. Plus, models sometimes "memorize" original works and reproduce them in their output (so-called memorization issues).

Interim Rulings 2024-2026:

• Germany, LAION v. Robert Kneschke (September 2024). The Hamburg court ruled for LAION: using protected images to build a scientific dataset falls under German § 60d UrhG (text and data mining for scientific research).
• USA, Bartz v. Anthropic (June 2025, interim ruling). The US District Court for the Northern District of California: using legally acquired books to train LLMs is fair use. But using pirated copies is infringement.
• Japan: A special amendment to the Copyright Act (2018, expanded in 2024) explicitly allows using protected works for AI training, as long as it doesn't harm the legitimate interests of copyright holders.

Globally, courts and regulators are leaning towards AI training being permissible under certain conditions: legal sources, no memorization, and opt-out options for rights holders. This cuts down existential risk for the AI industry, but doesn't close every issue.

Opt-out: How rights holders can block training on their data

In 2026, you've got a few technical ways to signal "don't use my content for AI training":

• robots.txt with User-agent for AI crawlers. Disallow for GPTBot, ClaudeBot, Google-Extended, CCBot, anthropic-ai, and others. Not legally binding, but major providers follow it.
• noai, noimageai metatags. Like noindex, but for AI. Supported by DeviantArt, Tumblr, ArtStation.
• C2PA training restriction. You can specify "do not train" in file metadata. Adobe supports it; Microsoft partially does.
• Platform opt-outs. Meta, X, and Reddit added an option in account settings to block AI training on your content.
• EU AI Act reservation. Under Article 53(1)(c) of the AI Act, rights holders can state a reservation of rights in machine-readable form — GPAI providers must comply.

Opt-out effectiveness is still limited. Models are already trained on petabytes of data collected before most of these mechanisms existed. But for future training, these signals are slowly gaining legal weight.

AI and Patents: A Different Story

Alongside copyright, there's a debate about patenting AI inventions. Here, the position is almost universal: only a human inventor can get a patent.

• US (Thaler v. Vidal, 2023): A federal appeals court upheld the USPTO's denial of a patent for an invention attributed to AI DABUS.
• UK (Thaler v. Comptroller, 2023): The Supreme Court confirmed the same approach.
• EU (EPO, T 1473/19): The European Patent Office rejected it based on the same logic.
• South Africa: The only jurisdiction where the DABUS patent was registered — but that was a procedural, not a substantive, case.

For businesses, this means using AI tools for development is standard practice. You can patent the results, but a human inventor who made a substantial conceptual contribution must be listed on the application.

Platforms and UGC: Who Owns Your AI Content

A separate issue is what happens to your AI images, videos, and texts once you publish them on platforms. This isn't about copyright itself, but rather the licenses platforms automatically grab.

• Instagram, Facebook: when you upload content, you grant Meta a non-exclusive, royalty-free, transferable, sub-licensable, worldwide license. That means Meta can use your AI content in its ad campaigns.
• TikTok: A similar broad license. Plus, TikTok can use your videos to train its own AI models.
• YouTube: The license is narrower, limited to how the service works. But YouTube banned content that «misleads viewers about real-world events» — that includes AI videos with realistic deepfakes.
• X (Twitter): Starting in 2024, a condition was added: Grok can train on your posts. You can opt out in settings.
• LinkedIn: It trains its own AI features on your posts. Opt out: Settings → Data Privacy → Data for Generative AI.

For commercial projects, this means: if you create unique AI content for paid use, pick your publishing platforms carefully. Free Instagram posts could end up as material for Meta's ad campaigns.

Moral Rights of the AI Content Author

In Continental European jurisdictions (France, Germany, Italy, Russia), they distinguish between economic and moral rights of an author. Economic rights are transferable; moral rights aren't. These moral rights include:

• Right of authorship (to be named as the author).
• Right to a name.
• Right to the integrity of the work.
• Right to withdraw the work from circulation.

For AI content, this creates an interesting situation. In Russia, under Article 1255 of the Civil Code, moral rights belong to a human author. If a human contributed significantly to creating an AI work, they get moral rights they can't transfer under any contract.

Practical consequence: even if you commissioned AI design from a freelancer and got full exclusive rights, the freelancer keeps the moral right to demand attribution as the author. This is rarely an issue in ad comms, but in the book industry, film, and journalism, it could lead to conflict.

What to Do, Practically

If You're Creating AI Content for Commercial Use

If You're Commissioning AI Content from a Contractor

When you contract with a freelancer or agency using AI, make sure to include:

• Output Rights Guarantee. The contractor confirms they used models with commercial rights and the output doesn't infringe on third-party rights.
• Significant Creative Contribution. The contractor confirms they made a significant creative contribution to the final result (this helps with future protection).
• Exclusive Rights Transfer. This is standard, but specify the contractor transfers rights to the extent they acquired them.
• Indemnification. The contractor indemnifies the client for losses in case of a third-party claim.

If You're a Platform

DMCA notice procedures (for US jurisdiction), transparent ToS, and a dataset opt-out option for rights holders are standard platform requirements by 2026. The EU DSA adds content moderation transparency obligations.

What's in Quantium ToS

The Quantium Terms of Service lays out three key provisions:

• Output Belongs to the User. All rights granted by the model provider (OpenAI, Google, BFL, Kuaishou, Stability) are fully transferred to the user.
• AI Content Labeling. Sora 2 videos include C2PA metadata. Gemini Image pictures have invisible SynthID. These are provider requirements, and we keep them.
• Responsibility for Use. Users are responsible for how they use the output: adhering to model ToS, third-party rights, and their jurisdiction's laws.

Quantium doesn't use your prompts for training. We only keep necessary data for billing and service operations (see Privacy Policy).

Related materials: Can You Sell AI Art?, Midjourney Alternatives, Full Terms of Service, Privacy Policy.

Read Also

Selling AI Art: Legal Nuances

16 min

The Future of AI

14 min

Marketer Case Study: AI in Agency Work

10 min