Microsoft will restrict the use of the facial recognition system


On June 21, Windows published a document called the Responsible AI Standard, in which it promised to minimize any harm caused by its machine learning software. The company promised that it would ensure compliance with stricter rules for using the software.

Microsoft will restrict access to AI tools and facial recognition models for predicting emotions, gender and age from images. From June 21, new customers will need to apply for access to the use of the facial recognition system in the Azure Face, Computer Vision and Video Indexer interfaces, and existing users will need to apply and get permission for permanent access to the platforms within a year.

From June 30, 2023, customers will not be able to access facial recognition features if their application is not approved. According to Microsoft experts, restricting access will add an additional level of control over facial recognition services.

Microsoft will also stop the possibility of facial analysis to determine emotional states and facial details (gender, age, smile, facial hair and makeup) from June 21. With this function, it is easier to create stereotypes and discriminate against a person.

Microsoft will continue to provide access to some of the technologies through integration into certain services for people with disabilities (for example, the Seeing AI application). Microsoft recommended that customers use the open-source Fairlearn project and the Microsoft Fairness Dashboard to understand the work of Azure Cognitive Services.

Start a discussion …