LEARNER · GLOBAL
Why spotting AI-written text is becoming nearly impossible
Modern LLMs (large language models—systems trained on billions of text samples) now produce prose so indistinguishable from human writing that detection tools are failing. FT reports the gap between AI and human authorship is closing fast.
WHY IT MATTERS
BFSI compliance and fraud teams must assume written submissions (client onboarding, regulatory filings, legal docs) may be AI-authored. Signature verification, document hashing, and provenance chains become critical controls.
Source: Financial Times · 2026-08-29