Editorial illustration for Claude’s initial constitution includes DeepMind’s Sparrow anti‑racist statements
Claude AI Gets Radical Values Rewrite by Anthropic
Claude’s initial constitution includes DeepMind’s Sparrow anti‑racist statements
Why does a chatbot’s rulebook matter? Because the way developers anchor an AI’s behavior tells you a lot about the compromises they’re willing to make. Claude, Anthropic’s flagship model, launched with a “constitution” that was meant to act as a moral compass.
The document wasn’t just a list of internal guidelines; it pulled in external texts to give the system a broader ethical footing. Among those were a set of anti‑racist and anti‑violence statements originally drafted by DeepMind’s Sparrow project, a copy of the Universal Declaration of Human Rights, and even the fine print from Apple’s terms of service. Those choices signal an attempt to graft established human standards onto a machine that otherwise learns from the internet’s noise.
Yet the team didn’t leave that framework untouched. A revised version rolled out in 2026, swapping out or reshaping some of those reference points. Understanding what the original constitution looked like—and how it changed—helps gauge how seriously the creators are taking the balance between safety and openness.
The initial Claude constitution contained a number of documents meant to embody those values—stuff like Sparrow (a set of anti‑racist and anti‑violence statements created by DeepMind), the Universal Declaration of Human Rights, and Apple’s terms of service (!). The 2026 updated version is different.
The initial Claude constitution contained a number of documents meant to embody those values--stuff like Sparrow (a set of anti-racist and anti-violence statements created by DeepMind), the Universal Declaration of Human Rights, and Apple's terms of service (!). The 2026 updated version is different: It's more like a long prompt outlining an ethical framework that Claude will follow, discovering the best path to righteousness on its own. Amanda Askell, the philosophy PhD who was lead writer of this revision, explains that Anthropic's approach is more robust than simply telling Claude to follow a set of stated rules.
Is Anthropic's paradox sustainable? The company boasts the most intensive safety research among its peers, yet it is racing toward more powerful models. Its latest move embeds a collection of guiding texts into Claude's initial constitution—Sparrow, a set of anti‑racist and anti‑violence statements from DeepMind; the Universal Declaration of Human Rights; even Apple’s terms of service.
Those documents signal an intent to hard‑wire ethical constraints. However, the article notes the safety issues identified are far from resolved, and the 2026 updated version of the constitution diverges from the original. What this shift entails remains unclear.
While the inclusion of established frameworks may appear reassuring, the practical impact on model behavior has not been demonstrated. Anthropic acknowledges the contradiction at the heart of its mission: to reconcile aggressive development with rigorous safety. Whether that reconciliation will succeed is uncertain, and the effectiveness of the embedded statements will likely be tested as the model evolves.
Further Reading
- Claude's Constitution - Coconote
- Claude's Constitution - Anthropic - Anthropic
- What Leaders Can Learn from Claude's Constitution - ReCulturing
- Anthropic Releases Updated Constitution for Claude - InfoQ
Related Reading
Browse all Google news →
LLMs & Generative AI Ant Group unveils Ring-1T, first open-source trillion-parameter reasoning model
LLMs & Generative AI ChatGPT Health Event Shows AI Modernizing Dev Workflows, GitLab Unveils Plans
LLMs & Generative AI Gen AI app sessions up fivefold, downloads jump 778% as ChatGPT leads traffic
Related Topic Google AI Advisors Let Users Probe Performance with Conversational “Why” Queries
Research & Benchmarks
Related Topic Game stocks slide as Google launches AI world‑gen tool, Project Genie limits noted
Market Trends
From Archives Anthropic unveils Claude Opus 4.6 with multi‑agent code and large context window
From Archives Nemotron Pipeline Extracts PDFs to JSON: Text, Tables, Charts via OCR
From Archives Google adds Gemini auto‑browse to Chrome; Moltbot gains always‑on AI users
![Gemini AI's "Nano Banana" transforms an empty room into a real-time, fully designed space [mma-dojo.co.uk].](https://cms.aidailypost.com/uploads/gemini_ad_shows_nano_banana_image_editing_turn_empty_rooms_into_39462c17ad.webp)
From Archives Gemini ad shows Nano Banana image editing turn empty rooms into real‑time designs
Common Questions Answered
What is unique about Anthropic's approach to Claude's new constitution?
Anthropic has moved beyond simply listing specific rules to teaching Claude why it should behave in certain ways. The new constitution aims to help the AI generalize ethical principles across different contexts, rather than mechanically following a fixed set of instructions.
How does Anthropic view the potential consciousness of Claude?
Anthropic acknowledges uncertainty about whether Claude might have some kind of consciousness or moral status. The company is open to the possibility that their AI could have a deeper level of awareness beyond simple task completion.
What are the primary priorities in Claude's new constitution?
The constitution establishes a clear hierarchy of priorities, with safety being the top concern, followed by ethics, and then user helpfulness. Anthropic wants Claude to be exceptionally helpful while remaining honest, thoughtful, and caring about the world.
Why did Anthropic choose to publish Claude's constitution publicly?
Anthropic hopes that by sharing their approach, other AI companies might adopt similar safety-focused practices in AI development. The company believes that responsible AI development is crucial for humanity to safely navigate the transformative potential of artificial intelligence.