When G7 industry, digital and technology ministers met in Montreal last week, the shift in tone was clear.

Artificial intelligence was no longer treated as a speculative technology or a distant ethical concern. It was framed, correctly, as a core economic and industrial force, alongside energy,

supply chains and trade. That alone marks real progress. The meeting produced some practical outcomes. Ministers emphasized AI adoption by small- and medium-sized enterprises, improvements to digital capability and stronger cooperation among trusted partners. For economies grappling with weak productivity growth, this turn toward implementation matters.

Yet for all its strengths, the Montreal meeting also revealed a growing gap between how governments are thinking about AI and how AI itself is evolving.

Much of the discussion remains anchored in an earlier phase of the technology. AI as a tool. Systems that automate tasks, generate content or assist decision making. These capabilities are valuable and governing them responsibly is essential. But they are no longer the frontier. The dominant trend is toward what might be called Identic AI — personal AI agents that function as digital extensions of ourselves.

We are entering an era in which AI systems do not simply respond to prompts but act on our behalf as extensions of each of us. They pursue goals, negotiate transactions and learn continuously from our behaviour. They increasingly resemble a digital self, persistent and empowered to operate across our daily work and personal lives.

They will “see you when you’re sleeping and know when you’re awake.” And, if managed well, will provide each of us with a superpower.

This shift fundamentally changes how leaders should approach policy.

Once AI becomes Identic, the central questions are no longer just about building infrastructure or regulating models, companies. They are about ownership, alignment and accountability. Who owns the personal agent? Whose interests does it serve? What authority has been delegated to it? And who is responsible when it acts?

To their credit, ministers spoke at length about trust and responsible deployment. But in a world of personal agents, trust becomes more intimate. People will not delegate real authority to AI unless they believe their agent is aligned with their values and accountable to them, not to a platform or distant institution.

This is where the conversation about sovereignty must evolve.

In Montreal, sovereignty was framed mainly in national and technical terms. Secure systems, reliable data and reduced dependence on foreign providers. These concerns are legitimate. But Identic AI introduces another dimension, one that is equally important. Personal sovereignty.

I was not in the room, but I am sure an elephant was there. The United States effectively owns the current generation of artificial intelligence. The companies that investors call the

Magnificent Seven now have a combined market value that is as large as the entire GDP of the other six G7 countries. When ministers discuss sovereignty, they are really discussing the sovereignty of those six. Canada is in that group. If the next era of AI involves the creation of digital selves, then we cannot allow those selves to be owned or governed by foreign platforms whose interests may not align with our own.

Recognizing this reality also forces a difficult conclusion. Regulation and domestic investment, although necessary, are not sufficient. If Identic AI is to serve individuals rather than institutions, then the entire AI stack must be rebuilt from the bottom up. It must shift from the centralization that defines today’s platforms to a decentralized model in which compute, data and identity are distributed across networks that no single company or state controls.

This is technically achievable and economically feasible, and early examples already exist. Without this structural shift, personal sovereignty over our digital agents will remain an aspiration rather than a right.

Without personal sovereignty, Identic AI risks concentrating power rather than distributing it. Personal agents could end up serving platforms instead of people, institutions instead of citizens or commercial incentives instead of human interests. The result could be efficiency without autonomy. Convenience at the cost of control.

Canada is well positioned to lead this next conversation. It combines world-class AI research with a strong tradition of public trust and inclusive institutions. Extending digital sovereignty from systems to people, from protecting infrastructure to empowering citizens and their personal agents, would place Canada at the forefront of the coming transformation.

Canada should also consider a national plan for what might be called Universal Basic Identic AI. If personal agents become essential tools for participation in the economy and in civic life, then access to them cannot be restricted to large corporations or wealthy individuals. A universal plan, developed in partnership with the private sector and civil society, would ensure that every Canadian has a trusted personal agent that reflects their values and remains under their control. This would broaden economic opportunity, strengthen autonomy and help prevent a new digital divide that is based not on connectivity but on agency itself.

  • Government-backed innovation cluster grants $128 million for AI projects
  • Brookfield, Qatar team up on US$20 billion AI venture

The G7 meeting in Montreal showed that governments are beginning to take AI seriously as an economic force. The next step is to take Identic AI seriously as a political one. As AI shifts from tools to personal representatives, sovereignty must extend beyond nations to individuals. The future of AI will be decided not only by who builds the systems, but by who controls the agents and whether they truly belong to us.

Don Tapscott’s new book, You to the Power of Two: Redefining Human Potential in the Age of Identic AI, co-authored by Joseph Bradley, was released last week.