PoliTraQ Blog

AI in the Public Service: Policy Windows for Ethics Vendors and Labour Groups

PoliTraQ

AI Meets Ottawa: How Responsible-Tech Firms & Unions Can Still Write the Rules

Intro:
Ottawa just hit “go” on a three-year sprint to bake artificial intelligence into every corner of federal work—without finishing the safety manual first. For responsible-AI vendors and federal unions, that gap is more than a risk; it’s a rare policy window where outsiders can still shape the standards that will govern thousands of civil-service jobs and millions of citizen interactions.

1. The 15-Second Version of Ottawa’s Plan

Treasury Board’s new AI Strategy for the Federal Public Service 2025-27 sells AI as a copy-and-paste cure for paperwork: automate the repetitive stuff so scientists, border officers and benefits clerks can focus on “complex and critical” files. Translation: departments are racing to stand up AI pilots before the next election cycle.

2. Three Governance Holes You Can Drive a Briefing Through

  • No public catalogue yet – Ottawa promised an open registry of every AI tool it uses; today only an internal list of ~300 projects exists.
  • Who’s in charge? – Titles like “chief data officer” and “chief privacy officer” still overlap, leaving sign-off authority murky.
  • No independent cop – The Professional Institute of the Public Service wants an external AI regulator; so far, crickets.

Each hole is a lobbying slot: demand disclosure templates, clear accountability lines and third-party audits before deployment scales.

3. Where to Speak Up (Dates Still in Ink)

  • TBS AI Centre of Expertise – being designed right now; reach out to help write the “project support” playbook.
  • Parliamentary HR Committee – reviewing workforce impacts; brief them on transition funds beyond the $50 M earmarked for creatives.
  • ESDC’s AI Adoption Strategy – open for sectoral feedback; unions can push for SWSP training dollars in their bargaining units.
  • New Minister of AI – fresh portfolio, looking for quick wins; responsible-tech vendors can pitch “made-in-Canada” solutions that check the ethics box.

4. Turn “Think AI” into “Think AI Responsibly”

The strategy explicitly says it won’t adopt AI “at all costs.” That single clause is rhetorical gold—use it to embed fairness tests, union consultation steps and Canadian-content procurement criteria inside every business case template.

5. Training Money Is on the Table

An “evergreen” civil-service curriculum is being written by the Civil Service Commission. Vendors can offer plug-and-play modules on bias audits; unions can insist modules include worker-rights clauses and re-skilling pathways.

Takeaway:
Ottawa’s AI locomotive is leaving the station, but the guardrails haven’t been bolted on yet. Whether you sell ethical toolkits or defend collective agreements, the next six-to-twelve months are your best shot to place transparency, labour protection and sovereign-tech requirements into the federal operating system—before scale makes them almost impossible to retrofit.