Police say they're assessing how to use emerging technology after report of 'risky' AI useNew Zealand police have been using generative AI tools like Whisper AI and Google Translate in criminal investigations, despite breaking internal safeguards and guidelines. Official documents reveal repeated violations of usage rules, including translating te reo Māori conversations and using unsecured platforms. While these tools helped speed up investigations, such as those related to the Auckland cult case, concerns arose about privacy and security risks. Police leadership acknowledged the risks but emphasized the importance of these technologies for handling large volumes of foreign-language materials. Although Whisper AI was temporarily suspended due to misuse, it was later reinstated. The police claim they now have approved solutions, though specific tools are not named, and their updated Technology Capabilities List no longer mentions translation services.
Bias read (Center): The article presents a balanced account of the situation, highlighting both the benefits of using AI in investigations and the associated risks and policy violations. It includes perspectives from police officials and external experts, providing a comprehensive view without overtly favoring one side
Why factuality (78): The article presents specific details about police use of AI tools like Whisper AI and Google Translate, citing official documents obtained through the OIA. It references statements from Detective Inspector Stuart Mills and legal expert Kathryn Dalziel, providing a clear account of the situation bas
Why objectivity (65): The article includes direct quotes from officials and experts, maintaining a relatively neutral tone overall. However, phrases like 'scary' and 'risky workarounds' suggest a critical stance toward the police actions, while the inclusion of Kathryn Dalziel’s quote labeling the behavior 'absolutely un
StuffIndependentCenterFactual: no official source document/info detectedObjective 805 days agoThe article discusses concerns around relying on artificial intelligence (AI) for medical advice, specifically highlighting the risks of trusting AI systems like 'Dr AI' with health-related decisions. It warns against over-reliance on such technologies, emphasizing that they may not be fully reliable for diagnosing or treating symptoms. The piece appears to caution against the potential dangers of allowing AI to play a central role in healthcare decision-making.
Bias read (Center): The article focuses on technology and healthcare, which are not inherently politically charged topics. There is no clear ideological framing or bias evident in the headline or content description provided. The focus is on general caution regarding AI use rather than any specific political issue or立场
Why factuality: no official source document/info detected
Why objectivity (80): The tone is cautious and warning-based but avoids overtly biased language. It does not favor one perspective over another and presents the issue as a general concern rather than taking a strong stance. Some phrases like 'our new overlords' suggest a slightly critical or skeptical view of AI, but thi
Police quietly set up AI lab to test new machinesNew Zealand Police established an AI innovation lab approximately one year ago to evaluate artificial intelligence applications for operational use, according to documents released under the Official Information Act. The initiative coincided with reports that officers had previously violated internal protocols regarding the unauthorized use of AI translation tools, raising concerns about data privacy and confidentiality. The lab has tested various technologies, including Meta’s Whisper and Seamless translation models, the forensic analyzer BelkaGPT, and an AI transcription application designed to expedite the processing of interview recordings. While the Serious Fraud Office utilizes similar AI tools for data triage and image categorization, the Police Force stated that all implementations undergo rigorous testing to meet legal and security standards. Recent internal communications indicate a shift toward lighter risk assessments for newer tools like Seamless, arguing that risks remain consistent with previously approved systems.
Bias read (Progressive): The article frames the story through the lens of accountability and transparency, highlighting internal rule-breaking and potential privacy risks ('serious and scary ways') rather than focusing solely on technological efficiency. By emphasizing the release of documents via the Official Information (
NZ
%2Ffile%2Fattachments%2F2995%2F2026-07-28_05-23-23_199117.jpeg&w=3840&q=75&output=webp&we)
