AI Has a Constitution Now. What Does That Mean for Us?

This week, Rob and Kev unpack one of the strangest and most revealing signals in AI right now.

Anthropic has published a “constitution” for Claude. Not internal policy. Not developer guidelines. A document written for the AI itself, explaining the values it is expected to understand and reason from.

That raises an uncomfortable question.

Why would an AI need a constitution at all?

In this episode, we explore:

• What it means when AI companies talk about “non-human entities”
• The tension between AI constitutions and emerging EU regulation
• Why agentic systems are changing how we work, not just what we build
• The shift from asking AI questions to assigning it work
• Whether this is a technical decision, a cultural signal, or both

This is not a debate about consciousness or sci-fi futures.

It’s about how language, mental models, and responsibility are already changing as AI moves from tool to actor.

If you’re working with AI, building products, or making decisions about how these systems are used in the real world, this conversation matters.

Hosted on Acast. See acast.com/privacy for more information.