The announcement comes as the US faces calls for police reform following the killing of a black man, George Floyd.
In a letter to the US Congress, IBM stated AI structures used in regulation enforcement wanted checking out "for bias".
One campaigner stated it was once a "cynical" cross from a company that has been instrumental in developing science for the police.
In his letter to Congress, IBM chief govt Arvind Krishna stated the "fight towards racism is as pressing as ever", placing out three areas the place the company desired to work with Congress: police reform, accountable use of technology, and broadening abilities and academic opportunities.
"IBM firmly opposes and will no longer condone the makes use of of any technology, consisting of facial attention science supplied via different vendors, for mass surveillance, racial profiling, violations of fundamental human rights and freedoms," he wrote.
"We agree with now is the time to start a countrywide communicate on whether or not and how facial consciousness technological know-how must be employed through home regulation enforcement agencies".
Instead of relying on probably biased facial recognition, the company entreated Congress to use technological know-how that would deliver "greater transparency", such as physique cameras on police officers and information analytics.
Data analytics is greater fundamental to IBM's enterprise than facial awareness products. It has additionally labored to enhance science for predictive policing, which has additionally criticised for viable bias.
'Let's now not be fooled'
Privacy International's Eva Blum-Dumontet stated the company had coined the time period "smart city".
"All round the world, they pushed a mannequin or urbanisation which relied on CCTV cameras and sensors processed via police forces, thanks to the clever policing structures IBM was once promoting them," she said.
"This is why is it is very cynical for IBM to now flip round and declare they prefer a country wide speak about the use of science in policing."
She added: "IBM are attempting to redeem themselves due to the fact they have been instrumental in creating the technical skills of the police thru the improvement of so-called clever policing techniques. But let's now not be fooled by using their modern move.
"First of all, their announcement used to be ambiguous. They speak about ending 'general purpose' facial recognition, which makes me suppose it will no longer be the stop of facial attention for IBM, it will simply be customised in the future."
The Algorithmic Justice League used to be one of the first activist agencies to point out that there have been racial biases in facial cognizance statistics sets.
A 2019 learn about carried out by means of the Massachusetts Institute of Technology located that none of the facial consciousness equipment from Microsoft, Amazon and IBM have been a hundred percent correct when it got here to recognising guys and ladies with darkish skin.
And a find out about from the US National Institute of Standards and Technology cautioned facial cognizance algorithms had been some distance much less correct at figuring out African-American and Asian faces in contrast with Caucasian ones.
Amazon, whose Rekognition software program is used through police departments in the US, is one of the largest gamers in the field, however there are additionally a host of smaller gamers such as Facewatch, which operates in the UK. Clearview AI, which has been instructed to cease the use of pictures from Facebook, Twitter and YouTube, additionally sells its software program to US police forces.
Maria Axente, AI ethics specialist at consultancy company PwC, said facial focus had established "significant moral risks, basically in improving present bias and discrimination".
He added: "In order to construct believe and remedy essential problems in society, motive as a great deal as income ought to be a key measure of performance."
