DE — heise online

Claude Code: Anthropic Tightens Rules Against Game Clones

AINovember 2, 20262 min read

Anthropic has added new limits to Claude Code, its AI coding assistant, so it will not copy specific existing games. Developers say the limits also block some of their own legitimate work.

New rules against game clones

Anthropic added two new instructions for its AI model Opus 5.5. They are meant to stop Claude Code from rebuilding specific existing games. Original games that follow well-known genres, or that are based on the user's own designs and materials, are still allowed. An Anthropic spokesperson confirmed the change. Anthropic developer Addy Osmani said the limit is based on the terms of use, which ban violating other people's intellectual property.

Why game clones became a problem

At the Steam Next Fest in October, AI-made game copies sparked debate. Some indie studios were afraid of copycats. They cancelled public demos or delayed playtests. Other developers reported blackmail attempts, where they were threatened with early release of copied games. Valve, which runs Steam, knows about the problem. However, it does not want to ban AI development tools completely.

Key facts

Two new instructions were added for Opus 5.5.
The change was confirmed to Game Developer by an Anthropic spokesperson.
Tom's Hardware collected several cases where Claude Code refused to help with developers' own projects.
Complaints include game mechanics, anti-cheat software, and reverse engineering.
One user said Claude blocked reverse engineering of LG smart TV software.

Blocked work and unclear causes

Developers say Claude Code refused some work on their own projects. It is not clear if these cases come from the new game rule or from other safety filters. Anthropic has not made a separate statement on the individual cases. The company has already fixed earlier safety filters for cybersecurity tasks that were too strict and blocked harmless requests. It says it will keep refining the game rule based on feedback.

Why it matters

The story shows a hard balance. Game makers want protection from copies made quickly with AI. But strict limits can also stop honest developers from doing normal work, such as reverse engineering. AI tools can also help rebuild old console games for PC and clone programs like Adobe Photoshop as open-source versions, though these are still limited. Developers who are blocked can try other AI services or local models with open weights.

Glossary

Claude Code: An AI coding assistant from Anthropic that helps people write and change software.
Anthropic: An AI company that makes the Claude AI models and Claude Code.
Steam: A large online store where people buy and download PC games. Valve runs it.

Vintage fun: Prodding AI agents using retro hardware just because you can

AWS developer advocate Darko Mesaroš connected a DECwriter teletype to AWS Kiro, a command-line agentic coding service, so the teletype can serve as a terminal for AI agents. He used a WiModem232 to emulate a Hayes-compatible modem for the serial connection and a Rust application to convert Kiro's output, obtained via Agent Client Protocol (ACP), into a format a teletype could handle. Mesaroš said the approach should work with any ACP-compatible coding assistant, though it lacks packet-drop handling and buffering.

AIVnExpressNov 2

Musk mocks Altman, says everyone deserves RAM too

Elon Musk mocked OpenAI CEO Sam Altman's remark that it is "bad" for users to run AI locally on their devices instead of reserving RAM for data centers. On November 2, Musk posted "Everyone deserves to own RAM" along with a meme featuring Christoph Waltz from Inglourious Basterds, and he made other jokes at Altman's expense. Altman did not respond to the posts.

AIZDNet KoreaNov 2

Hanmi Pharmaceutical Builds Domestic Chatbot Infrastructure for Digital Healthcare Platform 'eppiday'

Hanmi Pharmaceutical has signed a strategic MOU with FuriosaAI, Xperix, and its healthcare subsidiary Baegle Labs to move the AI chatbot of its digital healthcare platform 'eppiday' to a Korean LLM and domestic AI infrastructure. Under the agreement, Baegle Labs will handle app and chatbot development, Xperix will provide domestic AI servers and LLM inference, and FuriosaAI will supply domestic AI chips. The four companies have verified in a test environment that the AI coaching answer generation can run on domestic servers, and the commercial launch timing will be set after quality and safety checks.