Claude stopped helping with reverse engineering and game cloning
Developers report that Anthropic's Claude Code and/or the Opus 5.5 model are refusing even harmless reverse-engineering tasks, including work done through the popular Reverse Engineer Anything (REA) tool.
The cause may be a new text rule that allegedly tells the model to avoid "[recreating] a specific existing game." The restriction looks like an attempt to stem the flood of AI-cracked and cloned games. Some developers have recently been hit by extortionists who threaten to turn their demos into full games before the creator can finish and release the original.
According to Twitter user Chasm, the rule reads: "Do build original games, games in a genre, and games from the user’s own designs and assets. Don’t build a recreation of a specific existing game, even if it is renamed or called original."
There is no official confirmation that the text is genuine, but the wording looks plausible given the clear refusals developers have been running into over the past week.
The situation is tricky to analyze because the complaints vary so much. The rule does appear to exist, since the number of issues in the Claude Code GitHub repository keeps growing. Some of the authors may have had dishonest motives, but at least one of them appears to be a genuine developer with a real track record.
Reported refusals include the following:
Claude refuses to perform standard game work partway through development.
The AI won't replicate an individual game mechanic or patch a tabletop-style game.
According to Reddit posts, the model declined to analyze a simple Windows crash dump.
One user allegedly couldn't get its help removing junkware from a Samsung TV.
Some complaints involve a refusal to work with REA altogether.
Purported developers say Claude blocked anti-cheat development and security reviews for their own games, citing triggered safeguards.
The restrictions seem to have two motives. The first and most obvious is that Anthropic, along with other AI players, is protecting itself in advance from lawsuits by software companies and game studios over providing the means to reverse-engineer their programs. The second is tied to the wave of clones and extortion mentioned above.
There are no official statements from Anthropic or other AI companies so far. Anthropic isn't alone here, as OpenAI recently deactivated the account of a user who was trying to clean up the firmware on their TV.
- Claude solves problems better than candidates, and an Anthropic manager gave up coding interviews
- Japanese 3D artist used two AI models to create a room in Unreal Engine 5 without ever opening Blender
- Samsung TV owner used Claude AI to clear ads and tracking from his TV in 11 hours