Developers are downgrading their AI on purpose – and it's working

> We added two sentences to Claude Code’s prompt on Opus 5.5 to steer it away from recreating someone else’s specific game because using Claude to infringe IP is against our usage policy.

By using one of these LLMs you are already doing it.

The newest deepseek 4.1 will be replacing puff reading and parsing for me now at a fraction of the cost of claude models I was using before

- they ll just hop to the chinese models

- wonderful plan letting your competitor capture all the market share

“because using Claude to infringe IP is against our usage policy”

Claude is IP infringement incarnate, is it not?

> using Claude to infringe IP is against our usage policy

When I look up "hypocrisy" in the dictionary, this would be an excellent example.

Are people really surprised there's collateral damage from a hastily added couple of sentences ripe for misinterpretation? Did Anthropic measure the impact of the change at all?