AI textile tools face guardrail scrutiny

YORK – OpenAI’s disclosure last week of model misconduct raised concerns and questions at last week’s Cascale meeting in Athens over what guardrails may be needed as artificial intelligence is rolled out across supply chain software and data collection systems in the textile sector.

Questions raised at the Athens event were around AI guardrails: whether there should be a shared industry AI assurance protocol alongside established assessment methodologies? These might include “allowed use” policies, minimum traceability requirements, and AI accuracy benchmarks by task? Also disclosure templates – even independent audits?