AI ethics overhaul: Microsoft to phase out public access to controversial facial recognition tool

One of Microsoft’s AI-powered facial analysis products, which purports to determine emotions based on videos and images of participants, is apparently being phased out of public access. Because there are no universal facial expressions, and even if there were, they would vary between groups, and because it is blatantly unscientific to “equate exterior displays of emotion with inside feelings,” experts have long criticised such “emotion recognition” programmes.

Professor of psychology at Northeastern University, Lisa Feldman Barrett, told The Verge that while AI-powered emotion identification programmes may recognise frowns, “that’s not the same as recognising fury.” The topic of AI-based emotion recognition has been reviewed by Barrett.

The decision by Microsoft to discontinue this specific product is a part of a “bigger revision” of the organization’s AI ethics guidelines.

In order to stress accountability and “find out who utilises its services and stronger human control into where these technologies are employed,” the corporation has updated its Responsible AI Standards, which were first presented in 2019.

What does this mean, then?

Microsoft will first remove some parts of its facial recognition services from “Azure Face” and restrict access to others. Users will now need to apply to utilise Azure Face and provide specific information about their deployment plans to Microsoft. However, some use cases, such as the automated face blurring in pictures and videos, will continue to be available to everyone.

Microsoft is also retiring Azure Face’s ability to identify attributes like gender, smile, age, hair, facial hair, and makeup.

“Experts inside and outside the company have highlighted the lack of scientific consensus on the definition of ‘emotions,’ the challenges in how inferences generalise across use cases, regions, and demographics, and the heightened privacy concerns around this type of capability,” wrote Microsoft’s chief responsible AI officer, Natasha Crampton, in a blog.

According to the corporation, existing users will lose access to these capabilities starting on June 30, 2023, while new customers will no longer be able to take advantage of them starting on June 21.

The twist is now here. Microsoft is ending “public” access to these tools, but they will still be used in “at least one of its own products,” according to the company. For people who are blind, this product is an app called Seeing AI that “uses machine vision to describe the world.”


You can now write for and be a part of the community. Share your stories and opinions with us here.