Improving Audits for More Accountable Algorithms
June 24, 2022 | Grantee Publications | Technology in the Public Interest

Algorithmic audits can be a tool for accountability in potentially biased or harmful AI systems. But if they are inconsistent and unregulated, audits could make the problems worse. In “Who Audits the Auditors?”, the Algorithmic Justice League (AJL) shares the first field scan of the algorithmic auditing ecosystem. The scan unearthed best practices, tools and methods, and common barriers to effective auditing. MacArthur grantee AJL recommendations for companies, regulators and lawmakers, and auditors themselves. Increasing oversight, creating clearly defined standards, improving transparency, and engaging stakeholders in the audit process could help make them more effective.

 

Stay Informed
Sign up for periodic news updates and event invitations.

Connect with us on social media or view all of our social media content in one place.

The privacy of your data is important to us. We've updated our privacy policies in response to General Data Protection Regulation.