Police are no longer allowed to use Microsoft’s AI services

Small, well camouflaged and crucial: body cameras . Police officers are using them more and more frequently, and the cameras sometimes play a major role in investigations. The controversial technology gadgets not only record the officers’ immediate surroundings, but can also use technologies for real-time facial recognition – with artificial intelligence (AI). One service that makes this possible is Microsoft Azure OpenAI . But from now on the police are no longer allowed to use it for facial recognition, the US company reveals in its updated guidelines . This takes the bodycam debate into the next round.

Facial recognition with Azure OpenAI prohibited

Since the beginning of 2023, Azure OpenAI has provided access to generative AI models, including GPT-4 , Dall-E 3 or Whisper . Not only private individuals, but also members of the US police apparently benefited from this – and used the AI ​​service for facial recognition. Microsoft recently responded to this by changing its guidelines on May 2, 2024 . There, the company specifically prohibits such AI use “by or for a police agency in the United States.” However, this also applies worldwide to law enforcement authorities that want to identify or verify people using Azure OpenAI .

Police body cameras criticized

According to Microsoft, the revised specifications affect all AI models hosted on Azure OpenAI. In the future, it will be prohibited to use corresponding services for real-time facial recognition in body cameras or mobile dash cams . The US company leaves open why Microsoft is making this restriction. It is quite possible that the measure has legal reasons. Critics of AI facial recognition repeatedly raise concerns about the security, data protection and discrimination of the people recorded. A reaction from the authorities is still pending.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button