Microsoft stops selling emotion-reading technology and limits facial recognition

Microsoft said on Tuesday that it would stop selling technology that infers a person’s emotional state based on images of the face and would no longer provide unrestricted access to facial recognition technology.

The actions reflect efforts by tech companies to limit access to sensitive technologies after US and European lawmakers considered legal limits on them.

Since 2021, Microsoft has been evaluating whether emotion recognition systems are science-based.

“These efforts have created important privacy issues, a lack of consensus on the definition of emotions, and the inability to generalize a link between facial expression and emotional state across multiple cases, regions and demographics,” said Sarah Bird, director of artificial intelligence at Azure, Microsoft’s cloud computing arm.

Current customers of the company’s products targeting these technologies will have a year before they lose access to artificial intelligence tools that infer emotions, gender, age, facial hair, hair and makeup.

In 2021, Google Cloud took a similar approach and blocked 13 situations from its emotion-reading tool and put another four under review. The company is evaluating a new system to describe movements such as frowns and smiles, without trying to link them to emotions.

Microsoft said customers now need to get approval to use facial recognition services, which can allow users to register on websites or open doors by scanning their faces.