POV: Canada, AI, and the Case For a National Trust Layer

Chris McLellan
3 days ago
6 min read

Updated: 12 hours ago

This post is a companion piece to the presentation delivered to the CDO Council of Canada by Chris McLellan, Founder of the Ask AI nonprofit. It provides additional context and resources to support discussion on the role that the Data Collaboration Framework standard could play in the development of a 'national trust layer' for Canada.

Quick Summary

This article is a companion to a presentation made by Chris McLellan to the CDO Council of Canada and outlines why Canada needs a national trust layer for AI built on data control rather than data copies.

Trust in AI is eroding due to sidelined copyright, shadow integrations, undermined authorship, identity hijacking, automation outpacing upskilling, and deeply asymmetric AI ownership.
The “Big AI Squeeze” is accelerating data absorption across LLM models, LLM stores, SLMs, AI apps, RAG, agents, and ETLs, so trust will not improve without stronger access control.
A national trust layer would rest on data sovereignty, data residency, data localization, and data stewardship.
The Data Collaboration Framework, a Canadian and international standard, would offer a key complementary pillar by focusing on data minimization, reduction of data integration, data-level access controls, and data lifecycle management.
Research indicates that when people and organizations can grant and revoke access, they are more likely to participate in public good projects, including AI for good initiatives.

Bottom line: Public trust in AI and accelerated digital innovation would benefit by minimizing data in order to make meaningful access controls possible.

Problem: Trust Is Under Seige

The increased adoption of AI at home and work has accelerated a mistrust amogst Canadians with regards to the benefits of the technology:

"49% of Canadians expressed trust in their governments, slightly above the OECD average of 39%."

Government at a Glance 2025: Canada (OECD)

"Canada ranked sixth-lowest out of 47 countries in trust toward information generated by AI systems. 70% of Canadians believe AI will have positive outcomes, but 86% are concerned about loss of privacy."

Canada is lagging behind global peers in AI trust and literacy (KPMG, 2025)

"55% of Canadians aged 18-29 worry that AI and automation could soon force them to change careers"

Are We Ready? Canadians Voice Real Fears About AI and Work (Abacus Research, 2025)

"Today’s AI-equipped machines are pushing yields to new heights, but at what cost to a farmer’s privacy? As machines capture millions of data points in real time, it begs the question: What happens to all that data?"

Farm data privacy concerns grow with smart technology (Farm Progress)

"Most citizens (81%) recognise privacy as the most important factor to enhance trustworthy AI. Because AI typically relies on vast amounts of data, it raises concerns about potential misuse of such personal data used in the development and implementation of AI models."

OECD Privacy Guidelines: Review of the OECD Recommendation on cross-border co-operation and privacy 👇👇👇

Opportunity: The Positive Power of Control

There's ample research to support the notion that people, when provided with increased control and agency, are more likely to collaborate, knowing they can also revoke access to their contribution:

"When individuals have confidence in the protections surrounding their personal data, they are more likely to engage in online activities, share information, and participate in the digital economy."

Privacy and data protection (OECD)

"Clarity, transparency, and individual control over who has access to what data, when, and for how long are widely regarded as essential prerequisites for public data sharing support."

Patient and Public Willingness to Share Personal Health Data for Third-Party or Secondary Uses: Systematic Review (Journal of Medical Internet Research)

"What emerged from the data was the importance of an intrapersonal factor called "psychological ownership”, which appeared to foster the effortful and collaborative behaviors that allow cross-boundary work to proceed effectively."

Psychological ownership for overcoming departmental barriers to innovation: A Study of innovation handoffs (Journal of Engineering and Technology Management)

"When a person is autonomously motivated their performance, wellness, and engagement is heightened rather than if a person is told what to do (a.k.a. control motivation)"

Self-determination theory (Wikipedia)

"Ostrom’s work shows that when users are in control of their resources, they develop trust and, consequently, create, monitor, and enforce their own rules. This is known as self-governance or collective action."

Elinor Ostrom’s 8 rules for managing the commons

Innovation: The Data Collaboration Framework

A new Canadian and International standard that advocates data minimization as an effective path to increased control, IT efficiency, and TRUST in the systems and organizations that govern AI:

"As we move forward, it’s vital to recognize that advanced technologies now touch nearly every stage of our lives, from cradle to grave, and that managing who has access to our information will only get more difficult as AI tools and agents become part of how we live, learn, and work."

How Data Collaboration Can Support A National Trust Layer to Accelerate Responsible AI Innovation (Digital Governance Council)

"We don't think about it often, but the fact is that societies around the globe protect things of value, including property, currency, and IP, by making them difficult (and illegal) to copy. But when it comes to personal and senstitive data, it's basically been a free-for-all for decades."

Can Collaboration Between Humans and Artifical Intelligence Be Controlled? (Ask AI nonprofit)

Why Now? The AI Data Squeeze Is Accelerating

But while standards and regulations try to catch up, fast-emerging trends like LLM stores, enterprise SLMs, and agentic AI creating a surge in demand for personal and proprietary data:

"Investors say real-world data is the next critical layer in AI infrastructure."

Protege Raises $30 Million to Scale Real-World AI Data Access

"Here we lay out the position that small language models (SLMs) are sufficiently powerful, inherently more suitable, and necessarily more economical for many invocations in agentic systems, and are therefore the future of agentic AI"

Research: Small Language Models are the Future of Agentic AI (Cornell University)

"The future of AI assistants is not just about smarter agents, it’s about secure agents that can be governed and are built with an understanding of when not to act."

Why Moltbot (formerly Clawdbot) May Signal the Next AI Security Crisis (Palo Alto Networks)

Inspirations For Data Collaboration

Here are a few examples to inspire new data projects, based on the principles of Data Collaboration Framework, where control and collaboration relace copies and choas:

"The UN Quality of Life initiative aims to put people at the heart of sustainable urban development, ensuring that cities grow in ways that improve lives, foster well-being and leave no one behind."

UN Quality of Life Initiative

"The goal is to give individuals and families practical information to help guide life decisions, whether choosing where to live, work, retire or visit, while also giving city leaders a clearer picture of where investment and reform were needed."

Saudi Arabia, UN-Habitat unveil Quality of Life Index at WEFx

"GDPR's data minimization and purpose limitations create an increasingly complicated environment for cloud-based artificial intelligence in Europe for all enterprises. Small on-premises models can clearly provide compliance advantages: Data never leaves the enterprise boundary; complete audit trails remain within the enterprise boundary; third party processor risks are eliminated."

Small Language Models are Closing the Gap on Large Models

"All of the agent’s access is strictly pass-through, meaning users can only query tables they already have permission to access. When access is missing, it flags this or falls back to alternative datasets the user is authorized to use."

Inside OpenAI’s in-house data agent (OpenAI)

Data Collabration Framework: Minimize Data To Make Control Possible

Thanks to the dedicated work of the Digital Governace Council, the Data Collaboration Framework has is available globally on a free-to-download basis:

Standards Council of Canada

International Standards Organization

About Ask AI

Founded in 2017, Ask AI is a volunteer-run nonprofit whose mission is to help people navigate the opportunities and challenges associated with the increased adoption of artificial intelligence at work. As an independent organization, we provide a trusted, balanced perspective on this transformative technology that is reshaping industries and professions worldwide: