Alexa Fairness in AI - Amazon Research Awards (ARA) Call for Proposals

About this program

Alexa AI–Natural Understanding (via the Amazon Research Awards program) seeks proposals to fund research that improves fairness, transparency, explainability and accountability in AI, with an emphasis on broadly accessible conversational AI capabilities. Funded topics include theories of algorithmic fairness and trustworthiness, detection and mitigation of data and algorithmic biases, fairness-aware algorithm design, and metrics/methods for evaluating interventions (including human–machine collaboration) especially applied to NLU, NLG, speech/speaker recognition and computer vision. Selected principal investigators may receive one-year unrestricted awards (typically up to $80,000 USD) plus AWS promotional credits (up to $40,000 USD) and training resources; budgets must be specified in USD and exclude administrative overhead. Applicants must follow ARA program rules, submit proposals per the provided template (recommended ≤4 pages plus appendices), state whether they will make code/data open source, and agree to acknowledge support and report publications and outputs.