Microsoft on March 30, 2026 announced a new Critique feature for Microsoft 365 Copilot that uses Anthropic’s Claude to review and correct responses generated by OpenAI’s GPT inside Copilot Researcher. Microsoft says the multi-model workflow improved performance by 13.8% on the DRACO benchmark and also introduced Copilot Cowork in early access via its Frontier program. The update aims to raise answer accuracy and enterprise trust.
Scoring Rationale
Official Microsoft feature rollout combining Anthropic Claude and OpenAI GPT delivers significant novelty and industry-wide scope, backed by a 13.8% DRACO improvement. Credibility is high due to company announcement; actionability is strong for enterprise customers but adoption remains partial, so score reflects high impact with modest adoption risk.
Practice with real SaaS & B2B data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all SaaS & B2B problems
