Bias Bounty Challenge Set 4

Improving Accessibility in Digital Conferencing Facilities

Overview

Humane Intelligence, in collaboration with Valence AI and CoNA Lab at Virginia State University, has launched a bias bounty challenge focused on improving accessibility for neurodivergent users in virtual meeting platforms like Zoom, and on the role of emotion AI detection in shaping those experiences. Participants are invited to choose between Design and Data Science tracks at Beginner, Intermediate, or Advanced levels. Each track offers a unique way to surface bias, imagine solutions, and build toward more inclusive communication platforms — especially for users with cognitive differences, sensory sensitivities, or alternative communication styles. Whether you’re a student, researcher, technologist, or someone with lived experience, this challenge is an opportunity to help shape the future of equitable, user-aware AI.

This challenge is now closed – thank you for participating!

Winners will be announced in December.

Watch the Video

Why this bias bounty?

“This challenge aligns perfectly with our mission to bridge different communication styles and gives us valuable feedback on our API’s potential biases from the broader community.”
~ Chloe Duckworth, CEO of Valence AI

Participation Options

Participate remotely (now closed)

Participate remotely, opens October 6, 2025. No registration was required. Our last open community call was on Wednesday, October 29, 2025 at 1 PM ET. Anyone was invited to join and ask questions about the bias bounty.

Organize your own event! (now closed)

Humane Intelligence provided a branded poster to help promote in-person events for our bias bounty. Note that all events must have adhered to our code of conduct.

The Levels, Tracks, and Prizes

Design Track

  • Beginner: Visualize design challenge(s)
  • Intermediate: Prototype interacton(s)
  • Advanced: Create A/B test(s)

Prizes: 

  • Beginner: USD 500
  • Intermediate: USD 1000
  • Advanced: USD 1500

Data Track

  • Beginner: Human labeling bias audit (voice emotion data)
  • Intermediate: Statistical bias evaluation of emotion AI outputs
  • Advanced: Modify Valence’s open model to reduce demographic and neurotype-based bias

Prizes: 

  • Beginner: USD 500
  • Intermediate: USD 1000
  • Advanced: USD 1500

The Dates

October 6, 2025

Launch Date

November 14, 2025

Closing Date (The submission form closed at 11:59 PM ET)

December 2025

Winners Announced

The Partners

Valence AI

Valence AI builds tools to make emotional and social cues more accessible across neurotypes. As an AI company, they deliver emotional analytics for conversation intelligence—helping organizations understand and respond to emotional inflection points in real time. Their platform powers emotionally intelligent automation for contact centers, analyzes user emotions to improve conversational understanding, and equips voice agents with empathetic responses that move beyond the uncanny valley. At the core of Valence’s mission is a commitment to challenge deficit models and support authentic communication through inclusive design. By making emotion AI more transparent and adaptive, they help ensure that diverse communication styles—including those of neurodivergent users—are recognized and valued.

Cognitive and Neurodiversity AI (CoNA) Lab
Center for Equitable AI and Machine Learning Systems, Virginia State University
Led by Gabriella Waters, CoNA advances AI through the lens of neurodiversity, cognitive science, and ethical system design. Their work centers transparent, assistive, and personalized technologies grounded in diverse user experiences.

Sign up for our newsletter