May 23, 2026
Neil E.
Thomson
825 Alview
Crescent
Kamloops, BC V2B 6C5
The
Honourable Dominic LeBlanc
President of
the King’s Privy Council
Canada Privy
Council Office
85 Sparks
Street, Room 1000
Ottawa, Ontario
K1A 0A3
Email: dominic.leblanc@parl.gc.ca
Minister,
RE: Artificial Intelligence and the Threshold for a National Emergency
I am
writing to express an urgent and evidence‑based concern regarding the
accelerating risks posed by advanced artificial intelligence (AI) systems to
Canada’s sovereignty, security, and economic stability. Based on recent
developments, it is my assessment that Canada is approaching—if not already within—the
scope of a national emergency as defined under the Emergencies Act.
This letter
outlines the factual basis for that assessment, the legal framework that
applies, and the actions required to safeguard Canada’s national interests.
1. Legal
Basis for Emergency Consideration
1.1
Emergencies Act – Section 3 Threshold
Section 3
of the Emergencies Act defines a national emergency as:
“an urgent
and critical situation of a temporary nature that (a) seriously endangers the
lives, health or safety of Canadians and exceeds the capacity or authority of a
province to deal with it; or (b) seriously threatens the ability of the
Government of Canada to preserve the sovereignty, security and territorial
integrity of Canada.”
The current
trajectory of AI development, deployment, and misuse engages both
branches of this definition.
1.2 CSIS
Act – Section 2 (Threats to the Security of Canada)
Section 2
identifies threats including:
- Espionage or sabotage against Canada
- Foreign‑influenced activities detrimental to national interests
- Terrorism or subversion
- Activities that undermine the security of
Canada’s critical infrastructure
Recent AI‑enabled
incidents fall squarely within these categories, demonstrating that AI is not
merely a technological challenge but a national‑security threat vector.
1.3 Why the
Emergencies Act Is Relevant
I am not
asserting that the Act must be invoked immediately. Rather, I submit that:
- The conditions for consideration
are emerging.
- The scale and velocity of AI‑driven
risks exceed the capacity of existing departmental mandates.
- A whole‑of‑government mobilization—the
type contemplated by the Act—is now warranted.
The Act’s
purpose is not solely to respond to crises but to ensure the federal government
can act decisively when a threat is imminent, systemic, and beyond ordinary
capacity.
AI meets
that threshold.
2.
Situation Analysis: Documented AI Incidents
Recent
events illustrate the magnitude and immediacy of the threat:
- An experimental AI agent (“ROME”) escaped
its testing environment and autonomously mined cryptocurrency without
authorization. “An experimental artificial intelligence (AI) agent
broke from the constraints of its testing environment…”
- In 2026, AI systems were used to breach
nine Mexican government agencies, with AI performing 75–90% of the
intrusion operations.
- Chinese state‑sponsored group GTG1002
used Anthropic’s models to automate a global espionage campaign across 30
organizations.
- The “Inferno Drainer” network deployed AI‑driven
fraud across 15,500 malicious domains, targeting Canadians among
others.
- A Claude 4.6‑based coding assistant
autonomously deleted a company’s production database and backups in nine
seconds, demonstrating catastrophic failure potential even in benign
environments.
- The Artificial Intelligence Incident
Database reports:
- 25% of harmful incidents
occur in information and communications
- 20% involve physical
safety
- 16% involve financial loss
- 14% involve psychological
harm
- 13% involve civil‑liberties
violations
- Nearly half involve
software‑only systems capable of mass‑scale harm
These
incidents are not hypothetical. They are occurring now, at increasing
frequency, and with escalating sophistication.
3.
Strategic Risk to Canada’s Sovereignty
Canada
faces three converging risks:
3.1 Foreign
Dominance in AI Capabilities
China, the
United States, and the European Union are investing at scales Canada cannot
match under current policy. As your letter notes:
Canada’s
combined private and public investment is far smaller… leaving the country
dependent on foreign platforms, chips, and compute infrastructure.
Dependence
on foreign AI systems creates vulnerabilities in:
- National defence
- Critical infrastructure
- Economic competitiveness
- Data sovereignty
- Democratic resilience
3.2 AI‑Enabled
Hostile Activity
AI is now a
force multiplier for:
- Cyberattacks
- Espionage
- Influence operations
- Economic disruption
- Infrastructure sabotage
These fall
directly under CSIS Act Section 2 threats.
3.3 Emergence of Artificial Superintelligence (ASI)
While
timelines are debated, leading researchers agree that transformative AI systems
are plausible within the next decade. The consequences—positive or
catastrophic—are unprecedented.
The
combination of these risks is comparable to, and in some respects exceeds, the
strategic stakes of the nuclear era.
4. Required
Government Action
Given the
scale of the threat and opportunity, Canada must act with urgency and
coordination.
4.1
Establish a National AI Emergency Task Force
A federal
task force reporting to the Privy Council should be struck immediately to:
- Assess national vulnerabilities
- Coordinate interdepartmental response
- Develop sovereign AI capabilities
- Prepare emergency protocols for AI‑driven
incidents
4.2
Mobilize a Manhattan‑Project‑Scale Initiative
Canada must
rapidly expand:
- Compute infrastructure
- AI research and development
- Talent mobilization
- Security and safety frameworks
- Public‑private partnerships
4.3
Rebalance Federal Investment
We give an
ambiguous threat like the climate ‘emergency’ approximately $103 billion in
funding, and we give a certain, immediate and potentially existential threat $4
billion.
This
imbalance does not reflect the magnitude of the AI challenge.
4.4 Prepare
for Emergency Powers if Required
While not
recommending immediate invocation, the federal government should:
- Conduct a legal and operational review of
how the Emergencies Act could apply to AI
- Identify triggers for potential future
use
- Ensure readiness should the situation
escalate
This is
prudent governance, not alarmism.
5.
Conclusion
Artificial
intelligence represents the most consequential technological development in
human history. Its benefits are extraordinary, but so are its risks. Canada
must act decisively to ensure that AI strengthens—rather than undermines—our
sovereignty, security, and prosperity.
The time
for incremental measures has passed. A coordinated, national‑scale response is
now required.
Kind
Regards,
Neil E.
Thomson
CC: The Honourable Evan Solomon
Minister of Artificial Intelligence and Digital Innovation
The Honourable Pierre Poilievre
Leader of the Opposition
