A competitor launches a product that does what yours does, in the same order, with screens that feel familiar. Your team is certain they copied you. Whether you can do anything about it depends on what they copied.
The short answer: copyright stops people from copying your code, your text, your images, and your videos. It doesn't stop them from building the same features in their own code. A clone is an infringement only to the extent it takes protected expression.
The rule: expression, not ideas
Copyright protects "original works of authorship." 17 U.S.C. § 102(a). The next subsection takes a large area back out: protection doesn't extend "to any idea, procedure, process, system, method of operation, concept, principle, or discovery." § 102(b).
The rule is older than software. In Baker v. Selden, 101 U.S. 99 (1879), the Supreme Court held that the copyright in a book explaining a bookkeeping system didn't give the author rights over the system. Anyone could use the method. Only the author's explanation of it was protected.
For an app, the idea side of that line includes what the product does, its feature list, its workflow, its business model, and its pricing structure.
What copyright protects in an app
- Source code and object code, as literal text.
- Non-literal elements of the code, such as its structure, sequence, and organization, to the extent they're expressive and not dictated by function.
- Screen displays, including original artwork, icons, illustrations, and layout choices that are creative and not standard.
- Text: help content, onboarding copy, course material, and documentation.
- Media: photographs, video, audio, and animation.
- Original selection and arrangement of otherwise unprotectable material, which gets thin protection. Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991).
How courts filter a software claim
The leading test comes from Computer Associates International, Inc. v. Altai, Inc., 982 F.2d 693 (2d Cir. 1992). It has three steps:
- Abstraction. Break the program into levels, from the code itself up to its general purpose.
- Filtration. At each level, remove what isn't protectable: ideas, elements dictated by efficiency, elements required by external factors such as hardware, compatibility requirements, and industry practice, and material from the public domain.
- Comparison. Compare what's left with the accused program.
After filtration, there's often less left than the owner expects. Three decisions show how much comes out:
- Menus and commands. In Lotus Development Corp. v. Borland International, Inc., 49 F.3d 807 (1st Cir. 1995), Borland copied the entire menu command hierarchy of the Lotus 1-2-3 spreadsheet so that users could switch without relearning. The First Circuit held the hierarchy was an uncopyrightable method of operation. The Supreme Court affirmed by an equally divided vote.
- Interface conventions. In Apple Computer, Inc. v. Microsoft Corp., 35 F.3d 1435 (9th Cir. 1994), most of the graphical interface elements Apple asserted were either licensed or unprotectable ideas such as overlapping windows and icons. For what remained, the court required "virtual identity," not just substantial similarity.
- APIs. In Google LLC v. Oracle America, Inc., 593 U.S. 1 (2021), the Supreme Court assumed the Java API declarations were copyrightable and held that Google's reuse of them to let programmers work in a new platform was fair use.
When a clone crosses the line
Courts do find infringement when the copying reaches expression.
Tetris Holding, LLC v. Xio Interactive, Inc., 863 F. Supp. 2d 394 (D.N.J. 2012) is the best-known example. The defendant admitted it had studied the original game and set out to replicate it, while writing its own code. The court accepted that the rules of the game weren't protectable. It still found infringement, because the clone copied the game's visual expression: the look of the pieces, the dimensions of the playing field, the way pieces appeared and moved, and other design choices that weren't dictated by the rules.
Independent code doesn't save a clone that copies the creative look. A similar feature set doesn't condemn one that doesn't.
Evidence that separates copying from competition
Because the engineering details matter, look for evidence of literal or near-literal taking:
- Identical code. Decompile or inspect what's shipped to the client: JavaScript bundles, mobile app binaries, API responses. Matching variable names, comments, dead code, or distinctive bugs are strong evidence.
- Your strings and assets. Identical error messages, typos, image files with the same hashes, or your tracking IDs left in place.
- Your text. Help articles, emails, and marketing copy with the same sentences.
- Access. A former employee or contractor now at the competitor, an account the competitor registered on your service, or scraping patterns in your logs.
- Seeded markers. Fictitious entries or harmless unique strings you planted in your data or content, showing up in theirs.
Preserve it before you send anything: dated screenshots, archived pages, downloaded binaries, and hashes.
What covers the gap
| What they took | Tool |
|---|---|
| Code, text, graphics, video | Copyright |
| Confidential source code, algorithms, roadmaps, customer data | Trade secret law, including the Defend Trade Secrets Act, 18 U.S.C. § 1836 |
| Anything, in breach of an NDA, employment agreement, or terms of service | Contract |
| Your name, logo, or a confusingly similar brand | Trademark |
| A distinctive, non-functional look that customers associate with you | Trade dress (hard to prove for software) |
| A novel technical method, or an ornamental screen design | Utility or design patent, if you applied in time |
| Features, workflow, and business model, independently built | Nothing. That's competition. |
Practical takeaways
- Identify exactly what was copied before deciding it was infringement.
- Register the code and the content before a dispute, so statutory damages and fees are available. See When to register a copyright.
- Treat source code as a trade secret, and use terms of service that prohibit scraping and reverse engineering.
- Own what you're asserting. See Who owns the code your developer wrote?
- Preserve evidence first, then decide between a takedown notice, a demand letter, and a lawsuit.
Sorting copied expression from copied ideas is where my software background is most useful. If you think a competitor crossed the line, I can help you work out what they took and what it's worth pursuing.
Primary authority
- 17 U.S.C. § 102.
- Baker v. Selden, 101 U.S. 99 (1879); Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991).
- Computer Associates International, Inc. v. Altai, Inc., 982 F.2d 693 (2d Cir. 1992).
- Lotus Development Corp. v. Borland International, Inc., 49 F.3d 807 (1st Cir. 1995), aff'd by an equally divided Court, 516 U.S. 233 (1996).
- Apple Computer, Inc. v. Microsoft Corp., 35 F.3d 1435 (9th Cir. 1994).
- Google LLC v. Oracle America, Inc., 593 U.S. 1 (2021).
- Tetris Holding, LLC v. Xio Interactive, Inc., 863 F. Supp. 2d 394 (D.N.J. 2012).
Related resources
- Legal protection of software
- Copyright in software interfaces after Google v. Oracle
- Proving copyright infringement
- Intellectual property considerations and protectable content in mobile apps
- Data scraping after hiQ v. LinkedIn
- Running a clean room software development project
This article provides general information and is not legal advice. Copyright Office practices and fees change, and the law varies by circuit; confirm current requirements at copyright.gov and consult qualified counsel about your situation.
