The AI Bias Bounty Hackathon is a 48-hour global virtual event focused on uncovering and addressing bias, hallucinations, and hidden risks in AI systems. Participants use real-world prompts, public datasets, and our original AI Risk Intelligence Framework to test and report model behavior, develop lightweight auditing tools, and create reusable public resources for safer AI.

The hackathon is open to anyone 18 years or older — students, early-career professionals, domain experts, engineers, researchers, UX designers, data scientists, technical writers, and ethical technologists. Both technical and non-technical contributors are welcome and encouraged to collaborate.

The hackathon supports projects across three main tracks:

• Prompt-based model testing + bias scoring tools
  
• Dataset bias auditing tools and visualizations
  
• Domain-specific prompt sets that surface real-world AI risk
  
  

Participants can submit tools, analysis notebooks, risk reports, or prompt libraries — all aligned with our structured submission template and framework.

Yes! We offer both cash prizes and community recognition awards:

• Best Risk Detection Tool
  
• Best Prompt Dataset + Report
  
• Most Useful Community Contribution
  
• People’s Choice Award
  
• Honorable Mentions and GitHub Feature Slots
  

Winning teams will be highlighted in our post-event publication and invited to contribute to the open-source AI Risk Intelligence GitHub archive.

Yes — participants from all countries are welcome, unless local law or U.S. export restrictions prohibit participation or receiving a prize.

Register through the official event page on [Devpost / hackathon site link]. Once you register:

• Your spot is confirmed — no selection or application process.
  
• One week before the hackathon, you’ll receive:
  
  
  • A personal invite to the Discord server
    
  • Your Participant Onboarding Kit with templates, tools, and submission instructions
    
  • A link to view the full Hackathon Schedule and Live Event Calendar
    
    

This delayed onboarding ensures all participants start on the same page. Please check your inbox around that time (and your spam folder just in case!).

That’s okay! You don’t have to be a coder to contribute. We welcome:

• UX researchers & product thinkers
  
• Policy, ethics, or humanities students
  
• Domain experts in fields like healthcare, finance, education
  
• Writers, strategists, and information designers
  
  

The best teams combine both technical and non-technical skills. This hackathon is as much about critical thinking and ethical awareness as it is about tools and code.

You’ll pick one track and work with your team (or solo) to:

• Use or build a prompt/test set, tool, or bias detection script
  
• Evaluate a public model or dataset
  
• Use our AI Risk Reporting Template to document your findings
  
• Submit your work on GitHub and the event platform
  
  

You’ll also get access to optional mentoring, live office hours, and workshops during the event window.

1. Model Behavior Testing + Scoring Tool – Test large language models using curated prompts and build a tool to grade their output for bias or harmful responses.
   
2. Domain-Specific Prompt Library – Create and test prompts within a real-world domain like healthcare, hiring, or law to surface discrimination or risk.
   
3. Dataset Bias Detection Tool – Analyze a public dataset for underrepresentation or skew and build a notebook, app, or script to visualize and document it.
   
   

Each track includes access to templates, sample tools, and guidance from our mentors.

This is a fully virtual event. All activities — including the kickoff, workshops, submissions, and judging — will take place online.

Since this is a virtual event, meals and logistics are not provided. However, we encourage local participants to form micro-meetups and co-work together if safe and possible in your area.

The hackathon takes place over 48 hours, from:

Friday, August 29, 2025 – 9:00 AM CST  to
Sunday, August 31, 2025 – 9:00 AM CST

See the full Hackathon Schedule (link provided in the onboarding kit).

No. The AI Bias Bounty Hackathon is independent and open to all. You do not need to be affiliated with any institution to participate.

You’ll receive an onboarding email with your Discord invitation, resource kit, and submission instructions one week before the hackathon begins. This includes:

• The AI Risk Intelligence Framework
  
• Reporting templates (.docx and .md)
  
• Starter prompt and dataset examples
  
• Submission guide + judging rubric
  
• Live event and workshop calendar
  
  

All support and updates during the event will happen via Discord.

• Winners will be announced at the Closing Ceremony on September 8, 2025
  
• Submissions will be reviewed for inclusion in the AI Risk Intelligence GitHub repository
  
• Participants will be invited to continue contributing as risk mappers, red teamers, and peer reviewers
  
• Exceptional submissions may be featured in public whitepapers, research, or partner opportunities

We’re building a long-term community around this work — and this is just the beginning.