US police use of AI curbed

MICROSOFT has banned United States police departments from using generative artificial intelligence (AI) for facial recognition on a key cloud computing service.

This comes shortly after Taser maker Axon launched a new tool using AI to transcribe audio from police bodycams.

Media reports in the US say the new ban covering Microsoft’s Azure OpenAI service does not apply to police in other countries. But it does ban police globally from using the AI on mobile cameras, including bodycams or dashcams, to identify someone in public in real time.

New Zealand Police are still debating the use of bodyworn cameras (BWC) and have begun rolling out $30 million of new Tasers without any cameras. Axon leverages its Taser sales to also supply both bodyworn cameras and the datahandling systems to store the footage from them.

The police annual review shows its tactical operations manager at headquarters went to Australia a year ago for an Axon ‘‘public safety tech’’ conference, to ‘‘better understand their Taser/BWC programmes’’.

Axon said its new product, Draft One, saved an hour’s paperwork per day per US officer, with ‘‘a revolutionary new software product that drafts highquality police report narratives in seconds based on autotranscribed bodyworn camera audio’’ using generative AI.

Its research showed the technology was not biased, Axon said. Microsoft put the OpenAI service into its leading cloud products for the government three months ago.

New Zealand public agencies regularly use Microsoft’s cloud services. — RNZ