Experts Call for More Inclusive AI to Serve Diverse Users

Create user-centric AI design that is accessible and inclusive of different users' needs, like screen reader capability and speech recognition for accents.

Build a diverse team of AI reviewers and decision-makers to reduce bias; currently 80% of AI professors are men and fewer than 20% of AI researchers at top tech companies are women.

Regularly audit AI data sets for biases and create accountability policies to ensure data promotes diversity, equity and inclusion over perpetuating outdated prejudices.

Actively collect and curate more diverse data to fuel the creation of inclusive, ethical AI systems that represent people as multi-dimensional.

Require AI ethics training focused on recognizing and reducing biases while promoting diversity and inclusivity in AI development.