7 min read
Claude's Watermarking Policy Extends to Code
Claude's invisible marks could improve provenance, but Anthropic has not yet shown that they leave generated code quality and optimisation untouched.
Claude's invisible marks could improve provenance, but Anthropic has not yet shown that they leave generated code quality and optimisation untouched.
An opinion piece on the US government gating frontier AI releases, the security argument, and the commercial cost of slowing models down.