AI model upgrade for AI-Assisted Application Review

We've upgraded the model used for Criteria Evaluation in AI-Assisted Application Review to OpenAI’s GPT-5.4 family of models. It produces higher accuracy, more grounded reasoning, and more logically sound outputs when evaluating applicants against criteria. The newer models also have a later training data cutoff date, so they may perform better on certain criteria that rely on current world context, like company-related information.