Tasmania Police AI report trial cuts writing time but adds errors
Tasmania Police found that an AI tool nearly halved the time required to write family violence reports but also introduced errors, according to ABC News. Draft One, made by Axon Enterprise, analyzes audio from body worn cameras and creates narratives for officers to review, edit and approve.
The first phase used almost 150 recordings from historical incidents. A second phase in Launceston during 2025 compared reports written by officers with reports created using AI assistance. The evaluation found that Draft One reports took 42 minutes to write and validate, with estimated annual savings of 160 weeks of work and about A$434,187 in wages.
Draft One sometimes confused names and roles, and in a small percentage of cases missed a threat, assault or risk factor. In one case, it incorrectly reported that a woman described previous threats against her life when she had said there were no threats. The tool performed better in simple interactions than in incidents involving several people. The evaluation recommended a phased rollout once clear policies on the ethical use of AI are in place, with mandatory human verification and supervisor oversight, and said funding would be needed. Tasmania Police says a potential rollout is under review.
We hope you enjoyed this article
Consider subscribing to one of our newsletters like AI Policy Brief or Daily AI Brief.
Also, consider following us on social media:
More from Regulation
Oct 22 Fifth Circuit Vacates Mississippi DEI Injunction Over AI Errors Oct 22 UAE Fraud Operation Identifies $39 Billion in Linked Assets Oct 22 Reservoir Sues Suno and Anthropic Over AI Training Oct 22 Singapore Commits S$45 Million to Manufacturing AI Adoption Oct 22 HHS Plans Sole Source Claude Contract With AnthropicAI Policy Brief
Weekly report on AI regulations, safety standards, government policies, and compliance requirements worldwide.
Whitepaper
Tensordyne Napier: What If One Rack Could Do the Work of Nine?
Tensordyne
This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.
Read moreYou may also like
Tasmania Supreme Court Issues Generative AI Guidelines
Australian AI Inquiry Backs Independent Safety Checks
Australia's legacy systems vulnerable to AI agents, cyber expert warns
Appian survey finds 60% of Australian workers caught AI errors
Anthropic to Revise Leaked AI Safety Bill After Backlash
Daily AI Brief: the AI news that matters, in your inbox.