PL&B interviews our Director of Privacy about algorithmic bias

September 27, 2019

The UK’s leading source of data protection news PL&B, interviewed Neal Cohen, our Director of Privacy on a recent project we are undertaking with the ICO on researching algorithmic bias. 

The article hits on some of the concepts and research areas Onfido and the ICO are exploring in order to measure and mitigate algorithmic bias in artificial intelligence technology, with a particular focus on racial and other data related bias effects in biometric facial recognition technology. 

Other topics covered include: how AI fits under a new data protection framework, and the challenges of using personal data (such as facial images and biometrics) to train machine learning models when the intent is not to uniquely identify an individual or produce an impact on that individual. It also looks at how Onfido intends to explore the explainability of its machine learning models as well as their robustness and adaptability.

Read the full article first published in PL&B UK (September issue) here. If you have any questions about our research with the ICO, please contact us here.

Previous Article
Not all IDs are created equal: how geography impacts identity fraud
Not all IDs are created equal: how geography impacts identity fraud

Read part two of our '2019 Global Fraud Index' blog series to learn which documents are most frequently tar...

Next Video
The role of biometrics and KYC in seamless onboarding
The role of biometrics and KYC in seamless onboarding

Effective onboarding is a balancing act. With user friction, identity proofing and effective KYC all fighti...