Recently, police departments in the United States have begun to experiment with using AI chatbots to generate reports, sparking concerns about the reliability and legal implications of this new technology.
According to ABC News, the Oklahoma City Police Department is testing an AI system called "Draft One," developed by Axon based on ChatGPT technology. This system can quickly generate reports from audio recordings.
Image source note: The image is AI-generated and authorized by the service provider Midjourney
Officer Matt Gilmore was impressed with the results, stating, "This is a better report than I could write, and it's 100% accurate. It flows better."
While supporters praise the technology for saving a significant amount of time, critics express concerns about potential errors and biases. Local activist Aurelius Francisco is worried, pointing out that the technology is provided by the same vendor that also supplies stun guns to the police, which he finds "quite unsettling." He is also concerned that automation could make it easier for the police to harass, surveil, and exert violence against community members.
Currently, Oklahoma City is only using this technology for small incident reports, whereas in other cities like Lafayette, Indiana, its use is less stringent. Police expert Andrew Ferguson advises caution, worrying that "the convenience of automation and technology could lead to less meticulous report writing by officers." He emphasizes that the accuracy of alerts is crucial for legal judgments.
Axon's CEO Rick Smith sees great potential in this technology, believing that officers choose this profession because they want to engage in actual police work, and spending half a day on data entry is undoubtedly tedious. He mentions that Draft One has received the best response among all Axon products, but also points out that officers still need to take responsibility for their reports.
Key Points:
🔍 The Oklahoma City Police Department is testing the AI chatbot "Draft One" for quickly generating reports.
⏳ Supporters praise it for saving time, but critics worry about potential errors and biases.
⚖️ There is currently no nationwide AI usage guidance, and experts call for caution in automating police alerts.