Claude's new system prompt really doesn't want to reproduce song lyrics
- ID
- 20702
- Status
- summarized
- Published
- 02 Sep 2026, 10:16 PM
- Fetched
- 02 Sep 2026, 11:23 PM
- Provider
- Simon Willison
- Category
- developer-ai
- Original URL
- https://simonwillison.net/2026/Sep/2/claudes-new-system-prompt/
- Source URL
- https://simonwillison.net/atom/everything/
Summary
- Score
- 6.5
- Created
- 02 Sep 2026, 11:23 PM
- Tags
- Audience
- ai_agent_usersai_ml_learnersdevelopers
What happened
Anthropic updated Claude's system prompt (Fable 5.1) to aggressively block reproducing song lyrics, poems, and book passages — including partial quotes, hooks, and line-by-line pasting — and to persistently decline reworded requests within a conversation. The change landed within days of Sony Music Publishing and Warner Chappell suing Anthropic for training on lyric databases. New rules also forbid generating copyrighted characters, logos, and artwork via code (SVG, canvas, CSS, ASCII art).
Why it matters
If you build products on Claude APIs or agents that handle user-generated creative content, expect more aggressive refusals on lyrics and copyrighted visual output — design your UX and fallback flows around this. The pre-1929 cutoff and Claude's self-assessed date uncertainty mean edge cases will fail silently, so test your prompts against the new guardrails before shipping.
Discussion angle
How lawsuit-driven guardrail changes in base model system prompts cascade into your agent pipelines — and whether you should pin to specific prompt versions or build your own refusal-handling layer.