Find a story

Search Spins

Search titles, summaries, and missing voices across published articles — press releases, announcements, and media coverage.

0 results for “text watermarking”

SPIN Processed News Frame: The Cushion

Anthropic says text watermarking scheme relies on inconsequential words - The Register

Anthropic disclosed that its text watermarking technique operates by subtly altering low-impact, functionally redundant words — not core semantic content — to embed detectable signals without affecting meaning.

Spin 65% Claim Present in Source AI Risk Moderate
The Register AI / Software via Google News

Aug 15, 2026

SPIN Processed News Frame: The Halo

How Claude's text watermarking works - Anthropic

Anthropic has implemented an invisible text watermark in Claude-generated outputs to signal AI origin, prompting user backlash and subscription cancellations.

Spin 87% Needs Evidence AI Risk Moderate
Google News: Anthropic

Aug 15, 2026

SPIN Processed News Frame: The Hype

Making Open-Source Text LLM Watermarks Durable Against Merging

Researchers propose 'Merge-Adversarial Training' to make watermarks embedded in open-source LLMs resistant to model merging—a common post-training modification that previously erased such watermarks.

Spin 70% Claim Present in Source AI Risk High
arXiv Computation and Language

Jul 24, 2026