AI Copyright Legal Cases Analysis: What Businesses Should Know

AI copyright disputes are now one of the most important legal developments for businesses using generative AI. Tools such as ChatGPT, Midjourney, DALL·E, Stable Diffusion and other AI systems are changing how businesses create text, images, videos, software, designs and marketing materials.

However, the legal risks are growing quickly.

Recent lawsuits involving AI companies, publishers, authors, media studios and content owners show that AI copyright law is still developing. The major issues include whether copyrighted works can be used to train AI models, whether AI outputs can infringe copyright, who owns AI-generated content, and what businesses should do before using AI-generated materials commercially.

For Singapore businesses, many of the headline cases are overseas cases, especially from the United States. They are not automatically Singapore law. However, they are still important because they show the direction of global disputes, the arguments being tested, and the risks businesses should consider when using AI-generated content.

This article analyses key AI copyright legal cases and explains what businesses should know.

Why AI Copyright Cases Matter for Businesses

AI copyright cases matter because businesses are now using AI in daily operations.

AI tools may be used to create:

• Blog articles

• Website copy

• Product descriptions

• Social media captions

• Advertisements

• Images

• Videos

• Presentations

• Software code

• Training materials

• Brand concepts

• Product mockups

• Business reports

These outputs may appear original, but legal questions remain.

Businesses need to consider:

• Whether AI tools were trained on copyrighted materials

• Whether the output resembles protected works

• Whether the business owns the output

• Whether the output can be protected by copyright

• Whether the output can be used commercially

• Whether confidential materials were uploaded into AI tools

• Whether human review and editing were performed

• Whether platform terms allow the intended use

If your business needs copyright advisory and digital rights management in Singapore

AI copyright risk should be reviewed as part of your digital content strategy.

The Two Main AI Copyright Issues

Most AI copyright cases focus on two main issues.

Issue 1: Input Risk

Input risk concerns whether AI companies or users used copyrighted works to train or prompt AI systems without permission.

This includes questions such as:

• Were books, articles, images, videos or music used for AI training?

• Were the works lawfully obtained?

• Were pirated copies used?

• Is AI training fair use or infringement?

• Does training compete with the original market?

• Does the AI system store or reproduce protected works?

Issue 2: Output Risk

Output risk concerns whether AI-generated outputs infringe copyright.

This includes questions such as:

• Does the AI output reproduce protected expression?

• Does the output resemble a copyrighted character, image, article or artwork?

• Does the output create a derivative work?

• Can users generate near-verbatim content?

• Are safeguards sufficient?

• Who is responsible for infringing outputs?

For businesses, both issues matter. Even if a business did not train the AI model, it may still face risk if it uses infringing outputs commercially.

For a broader guide, read AI generated content copyright Singapore

Case 1: Anthropic and AI Training on Books

One of the most important AI copyright decisions involved Anthropic.

In June 2025, a US federal judge ruled that Anthropic’s use of books to train its AI system could qualify as fair use under US copyright law. However, the judge also said that pirating authors’ books could not be justified. Reuters described the ruling as a key decision because fair use is a central defence for AI companies in copyright cases.

A later Reuters analysis of 2025 copyright law developments described the Anthropic case as drawing a line between fair use for lawfully acquired training data and a hard stop on piracy.

Business Takeaway from Anthropic

The business takeaway is not that all AI training is automatically legal.

Instead, the case suggests that courts may distinguish between:

• Lawfully acquired training materials

• Pirated or unlawfully obtained materials

• Transformative AI training arguments

• Market harm to copyright owners

• Storage and reproduction of source materials

• Output-based infringement

For Singapore businesses, the practical lesson is this: lawful access and documentation matter.

Businesses using AI systems should avoid uploading pirated, unauthorised or confidential materials into AI tools.

For copyright basics, read how copyright works in Singapore

Case 2: OpenAI and Author Copyright Claims

OpenAI has faced multiple copyright lawsuits from authors, publishers and content owners.

In October 2025, Reuters reported that a New York federal judge denied OpenAI’s early request to dismiss authors’ claims that text generated by ChatGPT infringed their copyrights.

In March 2026, Reuters also reported that Encyclopedia Britannica and Merriam-Webster sued OpenAI, alleging illegal use of copyrighted materials for AI training and that ChatGPT produced near-verbatim summaries of their content.

These cases show that AI copyright disputes are not only about training data. They also involve whether AI outputs reproduce or closely resemble protected materials.

Business Takeaway from OpenAI Cases

Businesses using ChatGPT or other text-generation tools should avoid treating AI output as risk-free.

Risks may arise where businesses ask AI tools to:

• Summarise paid articles

• Rewrite copyrighted articles

• Generate content based on proprietary reports

• Produce text in the style of a specific living author

• Create near-verbatim passages from protected materials

• Draft content based heavily on competitor websites

• Process documents they do not have permission to use

Businesses should use AI as a drafting assistant, not a copying engine.

If your team uses ChatGPT, read ChatGPT copyright risks

Case 3: Disney and Universal v Midjourney

AI image generation has produced some of the most visible copyright disputes.

In June 2025, Disney and Universal sued Midjourney in the United States. Reuters reported that the studios accused Midjourney’s AI image generator of copyright infringement involving well-known characters and called the platform a “bottomless pit of plagiarism.”

The dispute is significant because it focuses heavily on AI-generated images that allegedly reproduce or imitate famous protected characters.

For businesses, this matters because many marketing teams use AI image tools for social media, advertising, campaign visuals, product concepts and brand storytelling.

Business Takeaway from Disney and Universal v Midjourney

Businesses should avoid prompting AI tools to generate images resembling famous characters, films, franchises, artworks or branded worlds.

High-risk prompts may include requests to create images:

• Involving famous movie characters

• In the visual style of a known franchise

• Based on well-known animated characters

• Using recognisable brand elements

• Featuring protected costumes, logos or mascots

• Designed to look like an existing copyrighted scene

For a deeper guide, read Midjourney copyright issues for businesses in Singapore

Case 4: Warner Bros Discovery v Midjourney

In September 2025, Warner Bros Discovery also sued Midjourney.

Reuters reported that Warner Bros accused Midjourney of copyright infringement involving characters such as Superman, Batman, Wonder Woman, Scooby-Doo and Bugs Bunny. The lawsuit sought damages, profits and an injunction against further infringement.

This case reinforces the same risk: AI-generated images can create serious legal concerns when they reproduce or closely resemble famous protected characters.

Business Takeaway from Warner Bros v Midjourney

Businesses should treat AI-generated images with caution when used for:

• Paid advertising

• Merchandise

• Product packaging

• Brand mascots

• Campaign characters

• Event posters

• Social media campaigns

• Commercial videos

• Retail visuals

Even if an AI tool allows commercial use under its platform terms, that does not guarantee the output is free from third-party copyright claims.

For image-specific ownership issues, read who owns AI generated images in Singapore
🟥 BACKLINK: /blog/who-owns-ai-generated-images-singapore

Case 5: Disney, Universal and Warner Bros v MiniMax

AI copyright disputes are not limited to US-based AI companies.

In September 2025, Reuters reported that Disney, Universal and Warner Bros Discovery sued China’s MiniMax, alleging that its Hailuo AI image and video service was built from intellectual property stolen from the studios.

This case shows that AI copyright disputes are becoming global. Businesses using AI tools from different countries should not assume that platform location removes copyright risk.

Business Takeaway from MiniMax

Businesses should review AI tools before using outputs commercially.

Consider:

• Who provides the AI tool

• What platform terms apply

• Whether commercial use is allowed

• Whether outputs are public or private

• Whether prompts and outputs may be used for training

• Whether the platform has copyright safeguards

• Whether outputs resemble protected works

• Whether indemnities or protections are offered

For digital content protection, read how to protect digital content in Singapore

Case 6: Thomson Reuters v ROSS Intelligence

Thomson Reuters v ROSS Intelligence is another important AI-related copyright case.

A 2025 analysis noted that a Delaware federal court issued a major decision concerning the use of copyrighted legal content to train an AI-related legal research tool. The dispute involved Thomson Reuters’ Westlaw headnotes and Ross Intelligence’s use of those materials to train a competing AI-driven legal research system.

This case is important because it involved a commercial competitor using copyrighted content to develop a competing product.

Business Takeaway from Thomson Reuters v ROSS

This case highlights risk where copyrighted materials are used to build a competing business product.

Businesses should be cautious when using third-party content to train or improve:

• Search tools

• Recommendation engines

• Legal tools

• Research platforms

• Knowledge databases

• AI chatbots

• Content generation systems

• Internal AI models

• SaaS products

The risk may be higher where the AI system competes with the market for the original content.

What Singapore Businesses Should Learn From These Cases

Although many major AI copyright cases are overseas, Singapore businesses should still learn from them.

The key lessons are:

• AI copyright law is still developing

• Platform permission does not remove all legal risk

• Training data and output similarity are both important

• Pirated or unauthorised source materials create higher risk

• Famous characters and branded content are high-risk prompts

• AI-generated output should be reviewed before commercial use

• Human creativity and editing matter

• Contracts should address AI use

• Businesses should document prompts, edits and approvals

• AI content should be part of IP risk management

IPOS states that AI raises practical copyright issues for copyright owners, AI developers and AI users, and provides resources on AI-generated content and possible infringement.

Singapore Law Perspective on AI Copyright

Singapore copyright law protects certain original works, but AI raises difficult questions around authorship, originality and liability.

A Singapore-focused IP analysis notes that under Singapore law, an AI system cannot be named as an author of an authorial work because Singapore copyright law requires an author for authorial works.

This matters because businesses may not always be able to claim strong copyright protection over raw AI-generated outputs with minimal human involvement.

For businesses, the practical approach is:

• Add substantial human input

• Edit and refine AI outputs

• Keep prompt and edit records

• Avoid copying protected materials

• Review platform terms

• Use contracts to allocate ownership and liability

• Avoid using AI output as core brand assets without legal review

For fair use issues, read fair use Singapore explained

AI Copyright Risk for Marketing Teams

Marketing teams face high AI copyright risk because they frequently create public-facing content.

Risky uses include:

• AI images resembling famous characters

• AI-written articles based on competitor websites

• AI-generated ads inspired by known campaigns

• AI videos using copyrighted character concepts

• AI content using unlicensed images as prompts

• AI tools used to rewrite third-party articles

• AI captions based on copied content

• AI-generated product visuals that resemble known brands

Marketing teams should use AI to create original campaigns, not to imitate protected works.

For brand-related IP strategy, read AP Swatch colours, Royal Pop, trademark and IP protection

AI Copyright Risk for Software and Technology Businesses

Technology companies face a different set of risks.

They should consider:

• Whether training data was lawfully obtained

• Whether open-source licences apply

• Whether source code generated by AI is safe

• Whether internal documents were uploaded into AI tools

• Whether customer data was used

• Whether employees used AI tools without approval

• Whether AI outputs are incorporated into products

• Whether contracts address AI-generated code

• Whether competitors’ materials were used to train internal systems

For startups, this is especially important because investors and buyers may review IP ownership during due diligence.

For startup legal planning, read legal requirements for startups in Singapore

For M&A due diligence, read mergers and acquisitions Singapore process

AI Copyright Risk for Agencies and Freelancers

Agencies and freelancers using AI should be transparent and contractually protected.

They should consider:

• Whether clients allow AI use

• Whether AI use must be disclosed

• Who owns final deliverables

• Whether raw AI outputs are acceptable

• Whether third-party materials were used

• Whether prompts include confidential client information

• Whether outputs are reviewed for similarity

• Whether indemnities apply

• Whether platform terms permit commercial use

Business contracts should now include AI-related clauses.

For contract planning, read business contracts Singapore guide

AI Copyright Risk for Brand Owners

Brand owners should be concerned about both sides of AI copyright.

They may use AI themselves, but they may also be copied by others using AI tools.

Risks include:

• AI-generated lookalike logos

• Copycat brand visuals

• AI-generated images of protected mascots

• Fake product images

• Brand impersonation

• Unauthorised use of product designs

• AI-generated social media content resembling official content

• Marketplace listings using copied brand assets

For brand protection, read trademark registration Singapore

For enforcement, read trademark infringement Singapore

What Businesses Should Do Before Using AI Content Commercially

Before using AI-generated content in public or commercial settings, businesses should take practical steps.

Step 1: Review the AI Tool’s Terms

Check:

• Whether commercial use is allowed

• Whether outputs are private

• Whether the platform may reuse prompts

• Whether the platform provides indemnity

• Whether restrictions apply to company size

• Whether generated outputs can be licensed or sold

Step 2: Avoid Protected Works and Famous Characters

Do not intentionally prompt AI tools to imitate:

• Famous characters

• Known artists

• Branded campaigns

• Movie franchises

• Logos

• Mascots

• Copyrighted scenes

• Competitor materials

Step 3: Add Human Creativity

Businesses should:

• Edit AI outputs

• Add original input

• Use human judgment

• Combine AI output with original work

• Avoid publishing raw outputs for high-value assets

Step 4: Keep Records

Maintain records of:

• Prompts

• AI tool used

• Date of creation

• Human edits

• Source materials

• Internal approvals

• Final version

Step 5: Use Contracts

Contracts should address:

• Whether AI may be used

• Who owns final work

• Whether AI use must be disclosed

• Responsibility for infringement

• Confidentiality

• Indemnities

• Commercial use rights

Step 6: Review Before Publishing

Check whether the output:

• Resembles known works

• Contains protected characters

• Uses third-party branding

• Repeats source material too closely

• Contains inaccurate legal or factual claims

• Violates platform terms

AI Copyright Checklist for Businesses

Before using AI-generated content, review:

• Is the AI tool approved for commercial use?

• Are platform terms reviewed?

• Was any copyrighted source material uploaded?

• Was confidential information uploaded?

• Does the output resemble existing content?

• Was human editing added?

• Are prompts and edits documented?

• Are employees trained on AI use?

• Are agency contracts updated?

• Are copyright and trademark risks reviewed?

• Is the output suitable for public advertising?

• Is the output being used as a core brand asset?

• Are licensing rights clear?

• Is legal review needed?

For a broader legal planning guide, read business legal checklist Singapore

Common AI Copyright Mistakes Businesses Make

Businesses often make avoidable mistakes.

Common mistakes include:

• Assuming AI output is always safe

• Asking AI to rewrite copyrighted articles

• Generating images of famous characters

• Using raw AI output as a logo

• Uploading confidential client documents

• Ignoring platform terms

• Not checking similarity to existing works

• Failing to document human edits

• Not updating freelancer and agency contracts

• Treating US fair use rulings as Singapore law

• Publishing AI content without review

• Assuming attribution solves copyright risk

For broader mistakes, read common legal mistakes businesses make in Singapore

Why Work with Absolute IP

AI copyright cases are changing how businesses should think about content, technology and brand protection.

Absolute IP helps businesses with:

• AI copyright risk review

• Copyright advisory

• Digital rights management

• ChatGPT content risk assessment

• Midjourney and AI image risk review

• Trademark and brand protection

• IP clauses for business contracts

• Licensing and commercialisation agreements

• AI content policy support

• IP dispute strategy

If your business uses AI tools for content, images, software, marketing or product development, contact Absolute IP at [email protected] for practical legal guidance.

Conclusion

AI copyright legal cases show that generative AI is not risk-free.

The major disputes involving Anthropic, OpenAI, Midjourney, Disney, Universal, Warner Bros, MiniMax and Thomson Reuters show that courts are examining both training data and AI-generated outputs.

For businesses, the safest approach is to use AI responsibly. Avoid copying protected works, review outputs carefully, add human creativity, document the process, update contracts, and seek legal advice before using AI-generated content for high-value commercial purposes.

AI can be powerful, but businesses must treat it as part of an intellectual property risk strategy, not just a content shortcut.