Riseline AI Draft Talk to us

Use cases & checklists

What Canadian police services have publicly reported doing with AI, and starter checklists for adopting, questioning and reviewing AI tools.

Reported usesChecklists

Reported in Canada

Ten publicly reported uses, as of 2 November 2026. Each shows its status, and a tap reveals what is claimed versus the evidence seen. Statuses may have changed since the report.

RCMP (Alberta and B.C.)

Axon Draft One drafts reports from body-camera audio; the officer edits and signs off.

Pilot, due to end Aug 2026
Claimed vs. evidence seen

Claimed: RCMP says it is piloting the software to see whether it can decrease the time and resources needed to create reports.

Evidence seen: No Western Canada report had reached a courtroom as of June 2026, per the RCMP. We did not find pilot results. Legal experts raised accuracy and court-process concerns. [CBC / Canadian Press, June 2026]

Fredericton Police Force

Axon generative AI drafts reports for non-charge files such as missing persons and medical calls.

Pilot, expanded (as reported Apr 2025)
Claimed vs. evidence seen

Claimed: Force says it boosts effectiveness and enhances transparency and accountability. Members must review and sign off.

Evidence seen: No evaluation seen. Pilot was expected to conclude Dec 2025. [Blue Line / Canadian Press, April 2025]

Calgary Police Service

Real-time translation service developed by Axon.

Testing (as reported Jun 2026)
Claimed vs. evidence seen

Claimed: Not stated in the source.

Evidence seen: No evaluation seen. [CBC / Canadian Press, June 2026]

York and Peel Regional Police

Shared system for comparing lawfully obtained crime-scene images with mug shots.

Launched May 2024 (as reported)
Claimed vs. evidence seen

Claimed: York says it increases investigative capacity and reduces manual comparison time; a match alone is not grounds for arrest.

Evidence seen: No independent evaluation seen. York states the tool does not involve Clearview AI. [Blue Line / Canadian Press, April 2025]

Edmonton Police Service

Two in-house tools: audio transcription, and records management that links cases.

In development (as reported Apr 2025)
Claimed vs. evidence seen

Claimed: Not stated in the source.

Evidence seen: No evaluation seen. [Blue Line / Canadian Press, April 2025]

RCMP NCECC

AI to help find AI-generated abuse material and sift large volumes of seized data.

In use (RCMP, Apr 2025)
Claimed vs. evidence seen

Claimed: RCMP investigators say it is faster and reduces officers' exposure to the material.

Evidence seen: Service-reported; no independent evaluation seen. [Blue Line / Canadian Press, April 2025]

RCMP

AI to generate case-specific word lists to help attempt decryption of lawfully seized phones.

Prototype contracts awarded (Apr 2025)
Claimed vs. evidence seen

Claimed: Not stated beyond the RCMP's description of the goal.

Evidence seen: Two companies were tasked with prototypes via Innovative Solutions Canada, per the report. [Blue Line / Canadian Press, April 2025]

Lethbridge Police Service

In-house AI-enhanced CompStat and Intelligence Portal for crime analysis and deployment.

In use (Blue Line, 2025)
Claimed vs. evidence seen

Claimed: Service reports a 24 per cent reduction in crime severity, faster response times and higher case clearance.

Evidence seen: These are the service's own figures; we did not find an independent evaluation. [Blue Line, 2025]

Toronto Police Service

Facial recognition listed in the service's AI inventory as high risk.

Governed under TPSB policy
Claimed vs. evidence seen

Claimed: Not applicable: this entry shows a governance mechanism.

Evidence seen: The OHRC notes the inventory identifies one facial recognition technology as high risk. [OHRC on TPSB policy]

RCMP and Toronto Police

Clearview AI facial recognition.

Ended; reviewed by regulators
Claimed vs. evidence seen

Claimed: RCMP said it used Clearview in a limited way, mainly for child exploitation investigations.

Evidence seen: OPC found the RCMP did not satisfactorily account for most of its searches and contravened the Privacy Act. Toronto used it in 84 investigations. [OPC, 2021] [CBC News]

Appearing here is not an endorsement or criticism. To suggest a correction or addition, see the corrections policy.

Starter checklists

Practical starters drawn from the sources on this site. They are not official guidance or legal advice; your policy, privacy office and counsel decide what applies.

Before a service pilots or adopts an AI tool

Based on the joint privacy guidance and the Toronto Police Service Board policy. [Joint guidance, May 2022] [TPSB policy]

  • Is there lawful authority for this specific use, and who confirmed it?
  • Is it necessary and proportionate, with a specific, evidence-based reason?
  • Has a privacy impact assessment been done, with a plan to monitor and re-assess?
  • Has accuracy been tested, including for bias?
  • Who is accountable inside the service, and for the vendor?
  • Are officers trained, and do they know what they remain responsible for?
  • What data is collected, why, for how long, and how is it secured?
  • Will use be disclosed to the public, and to the Crown where it may influence charges?
  • Is there board or oversight review, and public engagement in proportion to the risk?
Questions for a vendor

Derived from the same sources, with the Clearview lesson in mind: a public body cannot collect personal information from a third party that collected it unlawfully. [OPC, 2021]

  • Where did the training data and any reference database come from, and was it collected lawfully in Canada?
  • What independent testing exists for accuracy and differences across groups?
  • Where is our data stored and processed, who can access it, and does it train the vendor's models?
  • What logs and audit records exist, and can we produce them for a court, board or privacy commissioner?
  • Can outputs be traced to source material so a user can verify them?
  • What happens to our data and records if we stop using the product?
  • Which of these answers can we put in writing in the contract?
Supervisor review of an AI-assisted report

Adapted from a US Police1 article, which calls a report "a credibility document". [Police1] Adapt to Canadian disclosure practice with your Crown and counsel.

  • Does the narrative match CAD times, body-camera footage and physical evidence?
  • Does it contain specific, observable facts, or generic conclusions such as "the subject was evasive"?
  • Has the writing style changed suddenly? If so, ask the officer to explain the incident in their own words.
  • Does it read more tidily than a real scene usually does?
  • Are the approved tool and permitted data defined in policy, and was it followed?
  • Has the officer read every line and can they stand behind every word?
Before an individual uses an AI tool

The Police1 article advises services to define approved tools and what data may be entered. [Police1]

  • Is this tool approved by my service for this task?
  • What information may I enter, and what is off limits?
  • Who reviews my work, and do I know what I must disclose?
  • If unsure, have I asked my supervisor or privacy office first?

Last checked 2 November 2026.