Microsoft withdraws large facial recognition database from the Internet

Microsoft quietly removed a facial recognition database from the Internet this week, reports Silicon Angle. The database, MS Celeb, at some point contained 10 million images of about 100,000 people. That would make it the largest facial recognition database.

The intention was that the database should contain only images of celebrities. However, a researcher, Adam Harvey, claims that there were also photographs in the database of journalists, artists, musicians, activists, politicians, writers and academics. Under the Creative Commons licence, these photos can be removed from the Internet, which means that people could end up in the database without knowing it.

Various tech experts state in The Financial Times that Microsoft may have violated European privacy legislation GDPR with the database. “They probably took it offline because their lawyers were concerned that they had no reason to process special data as faces according to Article 9 of the GDPR,” says Michael Veale, a technology policy researcher at the Alan Turing Institute.

The database has been used by various organisations to train artificial intelligence, including IBM, Panasonic, Nvidia, Alibaba and Microsoft itself. Two Chinese companies also used the software: Sensetime and Megvii. According to the New York Times, these companies can be linked to the use of facial recognition by the Chinese state. The government would have used the software to track ethnic minorities.

Academic purposes

In a statement to The Financial Times, Microsoft says that the website was intended for academic purposes. “However, the problem is that the database has not completely disappeared. It still exists in various repositories on GitHub and the hard disks of countless researchers.

Adam Harvey argues that the database will probably continue to be used in research projects all over the world.