Reaction
Opus 5.5's claimed sub-1% false-positive rate in AI-content detection
A user calls the rate impressive but says LLM-based detectors still struggle on their more challenging evaluation sets.
TLDR
A user says a result shows today's frontier LLMs are much better at detecting AI content than they used to be. They call Opus 5.5's sub-1% false-positive rate impressive, while noting that LLM-based detectors still struggle on their more challenging evaluation sets.
Combined views
—
3 Sources, first seen ago
