IBM stops facial recognition due to possible use for mass surveillance

IBM will discontinue its general facial recognition and analytics products because of its potential for mass surveillance and police violations of human rights. The company points out the dangers of bias.

IBM executive Arvind Krishna wrote in a letter to Congress that his company no longer provides facial recognition and analytics software and clarified to The Verge that it is no longer developing this technology either. IBM speaks of stopping general purpose facial recognition, i.e. for broad general application, leaving the door open for delivery of specific applications with facial recognition.

According to IBM, the time has come for a national dialogue in the US about the use of facial recognition by the police and, if so, how it should be done. At the same time, the company speaks out against the use of facial recognition by other suppliers if this technique is used for mass surveillance, ethnic profiling and the violation of human rights.

Technology can help the police in their task of protecting citizens, IBM says, but the danger is that techniques promote inequality and discrimination. As far as artificial intelligence is concerned, suppliers therefore need to ensure that it has been tested for bias, systematic anomalies and biases that can sneak into the technology and which can be amplified.

Concerns about the development of facial recognition and its use have been with tech companies for some time. Microsoft, Alphabet and Amazon have already called for the techniques to be regulated. The topic has come back on the agenda as law enforcement agencies were found to be using a ClearView service, which used social media images for a sizeable facial recognition database, in tandem with US criticism of police violations. .