Microsoft has become the latest US company to limit the use of its facial recognition technology by police.
The firm said it would not start sales to US police departments until the country approves national regulation of the technology, which critics say is racially biased and easy to abuse.
These followed widespread protests over police brutality and racial discrimination.
Amazon on Wednesday banned police from using its tech for one year, while IBM earlier said it would stop offering the technology for “mass surveillance or racial profiling”.
The American Civil Liberties Union has campaigned against such software for years, warning there is a danger it will be used to for widespread “suspicionless” surveillance.
“Microsoft, Amazon and IBM have finally started to take action. But we still have a long way to go to forever end the over-policing and surveillance of black and brown communities,” the organisation said in a statement.
It called on US lawmakers to order an immediate “pause” on law enforcement use of the technology.
Federal vs local
Firms have been under pressure in recent weeks to respond to the protests triggered by George Floyd’s death in police custody.
Microsoft president Brad Smith said at an event that the firm had not sold to police departments and would not start “until we have a national law in place, grounded in human rights, that will govern this technology”.
Microsoft first called for national regulation more than two years ago, warning that inaction could lead such services to “spread in ways that exacerbate societal issues”.
Companies tend to favour national rules, rather than be forced to deal with a patchwork of local laws.
However, there are some concerns that a national law could be a way to override stricter local regulation.
San Francisco, for example, has already banned facial recognition technology by its police and public agencies.