Most stories about cloned indie games start with the developer finding out after the fact. Mark Brown's starts with an email asking first. In a new Game Maker's Toolkit video, Brown, who makes the channel and also makes games, explains that two hobbyist developers wrote to him about a game they'd built with an AI coding tool, Claude Code, based closely on his word puzzle game Word Play. They wanted to know how he felt before they released it.
The video is Brown working out his answer in public. It doesn't end with a ruling. It ends with a question to his audience. That makes it a more useful thing to watch than most takes on AI and games, because it shows how tangled the question gets once it's your own game on the table.
The email, and the seven thoughts
Brown reads the developers' message at 1:05, then sets out his reactions as a run of separate "thoughts", some of which pull against each other.
The first is plain economics. From 4:36, Brown explains that Word Play still brings in money he relies on, and that a free or competing version of the same idea could eat into it. It's the least glamorous worry in the video and probably the most widely shared.
The second goes the other way. Games have always been built on other games, Brown says, and he admits to feeling flattered that someone took Word Play as a starting point. He draws a line, though, between a mod, which needs the original game, and a standalone clone that takes the design without paying its maker.
The third thought is history. Around 8:34, Brown points out that clones are nothing new, using the story of Threes and the copycats that followed it. What's changed, he argues, is speed: by his account of the developers' own logs, the AI tool produced a playable version of the core idea in about 12 minutes.
From there it gets personal. At 11:09 Brown sets out his objections to generative AI, and then, with some honesty, admits to a twinge of jealousy at how fast these tools can build what took him months. Around 12:49 he asks whether the clone, with its extra features, is actually better than his game, and lands on the view that adding more isn't the same as designing well. From 15:43 he rejects the idea that developers should avoid simple mechanics just because simple things are easy to copy.
The rule Anthropic added, and Valve's advice
Brown's last thought, at 16:53, turns to the toolmakers. He notes that Anthropic has since changed Claude Code to steer it away from recreating existing games, and criticises AI companies for adding guardrails only after the harm is done.
That change is real and recent. Game Developer reported on 30 October that Claude Code's rules gained two lines, one telling it to build original games, genre games and games from a user's own designs, and one telling it not to build "a recreation of a specific existing game, even if it is renamed or called original." Anthropic's Addy Osmani said the company had added the sentences "to steer it away from recreating someone else's specific game" because using Claude to infringe IP is against its usage policy. Game Developer also noted a limit: people can switch to tools without such rules.
Valve's view, for what it's worth to a developer in Brown's position, came out a few days earlier. Inven Global reported that at a developer event in Korea on 27 October, Valve's Tom Giardino said clones were a problem long before AI, which has simply sped things up. His advice was community and steady updates: "Rather than relying on a single mechanism that someone can easily copy, you must continue to grow your game by adding features, personality, and creativity that stem from your own perspective as a developer." Brown, as it happens, mentions Giardino's advice in his discussion of whether simple designs are too easy to copy.
Where the audience landed
Brown closes at 18:52 by asking viewers to vote and comment on what he should tell the developers. The early comments, posted within six hours of upload, lean hard one way. Many viewers say he shouldn't let the clone go ahead, and several make the point that asking permission was the rare polite case, with less scrupulous cloners unlikely to ask at all. A few find the gesture of asking worth something. Very few think the clone is fine.
The quiet detail worth holding onto is the email itself. Brown's dilemma only exists because two people decided to ask. The question his video leaves hanging is what happens to all the developers who never get that courtesy, and who find out the way most of them always have.
From the creators
Watch the video
Reporting notes
