EUResearch
SSRN: Friedl urges treating AI content detectability as regulated infrastructure
In an October 12, 2026 SSRN preprint, Paul Friedl argues that watermarking and other AI-detection methods should be regulated as infrastructure, because EU, California, India and Chinese detectability mandates face interdependent actors and technical complexity that current frameworks under-address when countering disinformation, fraud and deepfake abuse.
Key points
- Friedl posted the contribution on SSRN on October 12, 2026.SSRN
- He notes EU, California, India and China now require AI-generated content to be detectable.SSRN
- He says watermarking and related tools could counter disinformation, fraud and deepfake abuse.SSRN
- Technical complexity and interdependent actors make detectability hard to regulate well.SSRN
- He argues current frameworks under-treat detectability and should regulate it as infrastructure.SSRN