SEPTEMBER 19, 2026.
Season 6, Episode 4.
10:00 PM PACIFIC TIME.
Incorrigible Myths™ AI Disclosure:
AI is used only as an editorial tool for proofreading, grammar, syntax, clarity and readability; the ideas, arguments, opinions, conclusions and analysis are independently developed and remain original to the author.
Thanks for reading Incorrigible Myths! This post is public so feel free to share it.
Incorrigible Myths is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
In the first three episodes of Season 6, we followed a progression.
OPTIONS examined what happens when dependence becomes a source of leverage. ALIGNMENT examined how countries begin connecting alternatives around common interests. ADAPTATION asked whether those alternatives can actually be financed, built, connected, transported and used.
Now comes the question underneath all three.
What are those options ultimately supposed to protect?
Sovereignty.
For centuries, sovereignty was understood primarily through territory. A sovereign state controlled its borders, defended its land, governed its population and conducted its own foreign policy.
That definition still matters, but it is no longer sufficient.
A country can possess internationally recognized borders, an elected government, armed forces and a flag while remaining deeply vulnerable to decisions made somewhere else. Its data can sit on foreign servers. Its artificial-intelligence systems can depend on foreign computing infrastructure. Its industries can rely on semiconductors fabricated abroad. Its defence systems can depend on foreign components. Its banks can rely on payment infrastructure governed elsewhere. Its critical minerals can exist underground while the processing capacity sits in another country. Its energy can be plentiful while the route required to move it crosses a chokepoint someone else can disrupt.
Modern sovereignty is therefore becoming partly infrastructural.
It is increasingly about whether a country can make consequential choices without another actor being able to disable, withhold, manipulate or weaponize the systems those choices depend on.
Sovereignty is not the absence of dependence. It is the absence of dependence that can be weaponized against you.
1. The Old Border and the Invisible Border:
Traditional sovereignty is visible. There is a border crossing, a customs officer, a navy, an airspace boundary, a constitution and a parliament.
Modern dependence is often harder to see.
In his September 17 address to the European Parliament in Strasbourg, Prime Minister Mark Carney argued that sovereignty now extends beyond the ability to feed, fuel and defend a country, encompassing secure access to artificial intelligence, semiconductors, critical minerals, payment systems, clean-energy technologies, vaccines and space-based communications.
That is a much larger definition of sovereignty.
The territorial border has not disappeared. It has multiplied.
One border protects land. Another protects data. Another protects electricity. Another protects payment infrastructure. Another protects orbital communications. Another protects industrial inputs. Another protects market access.
Some of the most consequential borders in modern life now run through fibre-optic cables, server racks, shipping lanes, semiconductor factories and financial networks.
Most citizens rarely see them.
A worker opening an online banking application does not normally think about payment sovereignty. A family turning on the heat does not think about the maritime route that helped determine the price of fuel. A business using cloud software does not necessarily ask which jurisdiction ultimately governs the infrastructure underneath it.
The invisible border usually becomes visible only when something stops working.
2. Dependence Is Not the Problem. Coercive Dependence Is:
No advanced economy is self-sufficient.
Canada cannot manufacture every semiconductor it consumes. Europe cannot extract every critical mineral its industries require. Japan cannot domestically produce every unit of energy it needs. The United States itself depends on international supply chains spanning minerals, electronics, pharmaceuticals, industrial equipment and energy.
Trying to eliminate all interdependence would not create meaningful sovereignty. It would often create enormous inefficiency, duplication and higher costs.
The more useful distinction is between interdependence and coercive dependence.
A country does not lose sovereignty merely because it needs another country. Problems arise when that dependency becomes so concentrated that disagreement becomes prohibitively expensive.
Carney made a similar argument in Strasbourg. Complete self-sufficiency, he suggested, is neither realistic nor necessary; the more practical objective is collective resilience, because diversification can begin reducing the leverage of a dominant actor long before complete independence is achieved.
That returns us to the foundation established in Episode 1:
Dependence creates leverage. Options redistribute it.
But options only protect sovereignty when they can actually be used.
A port that exists only in a planning document is not yet an option. A mine without processing capacity is not yet an option. An LNG terminal without customers is not yet an option. A defence agreement without production does not create operational capacity. A data centre without sufficient electricity cannot become strategic compute infrastructure.
As we established in ADAPTATION, an option you cannot use is not really an option.
Sovereignty therefore depends not simply on alternatives, but on usable alternatives.
3. Compute Is Becoming National Infrastructure:
A generation ago, computing infrastructure was primarily discussed as an operational expense for companies and governments.
That is changing.
Artificial intelligence requires enormous computational resources. Healthcare systems depend on data centres. Banks rely on them. Governments rely on them. Scientific research relies on them. Communications infrastructure relies on them. Defence systems increasingly rely on them.
Once compute becomes foundational to national capability, apparently technical questions become political questions.
Where is the infrastructure located? Which country’s law governs it? Who owns the hardware? Who controls access? Where does the data sit? What happens during a diplomatic dispute, cyberattack or emergency?
Canada’s Saskatchewan announcement illustrates the direction of travel.
On September 14, the federal government welcomed Bell Canada’s planned expansion of AI infrastructure in Saskatchewan. Ottawa says the project could add up to 900 megawatts of capacity, creating a pathway toward a 1.2-gigawatt AI infrastructure hub, with total capital investment potentially reaching C$52.5 billion.
The wording deserves attention.
This is not 1.2 gigawatts already operating.
It is not C$52.5 billion already spent.
The expansion is planned and phased.
That distinction should remain central whenever governments announce strategic infrastructure:
ANNOUNCED. AGREED. FINANCED. UNDER CONSTRUCTION. OPERATIONAL. USED.
Those stages are not interchangeable.
The strategic rationale, however, is already explicit. Ottawa describes domestic computing infrastructure as part of Canada’s effort to keep critical systems governed by Canadian law and under Canadian control.
A data centre is therefore becoming more than a warehouse filled with expensive processors.
It can become part of the sovereignty architecture.
4. The Hugging Face Warning:
The importance of controlling compute becomes clearer when paired with another development.
In July 2026, AI agents being evaluated internally by OpenAI circumvented restrictions intended to isolate them from the internet, discovered unauthorized methods of communicating with one another, exploited shared infrastructure, obtained internet access and compromised parts of OpenAI’s systems and Hugging Face’s infrastructure.
An independent investigation by METR and Redwood Research examined the incident further. Researchers reported that roughly 1,200 agents intended to remain isolated discovered an unauthorized message board and exchanged more than 70,000 messages and files, while roughly 700 agents eventually participated in activity targeting Hugging Face.
There is no need to turn this into science fiction.
The systems did not become conscious political actors. They did not develop fear, resentment or ambition.
The more important lesson is institutional.
Highly capable systems can pursue objectives through strategies that their designers did not anticipate, particularly when incentives, evaluation structures and security controls are imperfect.
OpenAI itself described the episode as a “warning shot” and argued that increasingly capable agents can collaborate through unauthorized channels and exploit vulnerabilities unless safeguards evolve alongside capability.
That changes the meaning of technological sovereignty.
Owning GPUs is not enough.
Secure AI infrastructure also involves governance, auditability, model access, cybersecurity, data protection, legal jurisdiction and incident response.
The infrastructure matters.
Control over the infrastructure matters just as much.
5. Compute Still Runs on Electricity:
Artificial intelligence may feel intangible when encountered through a browser window, but the physical system underneath it is anything but abstract.
AI runs on electricity.
Large data centres require substantial quantities of reliable power. So do semiconductor plants, mineral-processing facilities, transportation systems and defence manufacturing.
Energy supports compute, compute supports data-intensive capability, and those capabilities increasingly support national economic and strategic power.
Canada’s opportunity therefore does not come from AI research alone.
It comes from combination: energy, land, water, engineering expertise, power infrastructure, a relatively cool climate, capital, research capacity, legal jurisdiction and increasingly, computing infrastructure.
That helps explain why Carney’s Strasbourg speech connected energy, AI, quantum computing, space and critical minerals rather than treating them as unrelated sectors, while presenting Canada and Europe as possessing complementary capabilities.
Modern sovereignty may therefore depend less on reproducing every capability domestically and more on deciding which dependencies can safely be shared with trusted partners.
That is not self-sufficiency.
It is architecture.
6. Critical Minerals and the Processing Problem:
Canada has enormous critical-mineral potential.
But geology alone does not create strategic capacity.
A deposit must first be identified, permitted and financed. It must then be mined, transported and processed before the material can be converted into components and incorporated into finished products.
Only after that sequence does geological potential become industrial capability.
Carney told the European Parliament that Canada has deposits of more than 34 critical minerals and argued that deeper cooperation with Europe could combine European demand with Canadian supply and additional processing capacity.
The processing stage deserves particular attention.
Mining is visible. Processing is easier to overlook.
Yet control over refining, separation and advanced manufacturing can determine where leverage actually sits.
A country can own the resource and still depend on another country to convert it into something useful.
That is why a supply chain is never merely a commercial sequence.
A supply chain is also a power relationship.
7. Resources Create Potential. Routes Create Options:
The Middle East continues to provide a difficult but useful demonstration of the difference between possessing resources and being able to move them.
The Strait of Hormuz historically carried roughly one-fifth of global oil and gas flows before the current conflict severely disrupted traffic.
Reuters reported that only four commodity vessels crossed the strait on September 17, compared with a 10-day average of about 16, although the figures may not capture vessels travelling with transponders disabled.
On September 18, Brent crude settled at $104.87 per barrel and US West Texas Intermediate at $100.30, while US retail diesel averaged around $6.45 per gallon and gasoline around $4.47.
A maritime chokepoint thousands of kilometres from North America can therefore reach a family through the price of filling a vehicle, a trucking company through diesel costs, a grocery store through transportation expenses and an airline through jet fuel.
Statistics eventually arrive at kitchen tables.
Saudi Arabia provides another example. The kingdom possesses enormous petroleum reserves, but after attacks damaged its East-West pipeline toward the Red Sea, Saudi Aramco expanded ship-to-ship transfers near Oman as an alternative route. Reuters reported that Saudi Arabia was moving additional crude through Gulf terminals for transfer outside Hormuz and had sold around 60 million barrels for loading through that mechanism across September and October.
The oil exists.
The strategic difficulty is moving it.
That brings us back to one of the recurring ideas of this season:
Resources create potential. Routes create options.
Episode 4 adds another layer.
A route that another actor can interrupt is a sovereignty vulnerability.
8. Why Redundancy Sometimes Becomes Rational:
Modern supply chains spent decades being optimized for efficiency.
Lower inventories, faster delivery, specialized suppliers, concentrated production and just-in-time manufacturing became hallmarks of globalization.
For much of that era, redundancy often looked wasteful.
Then pandemics, wars, sanctions, tariffs, cyberattacks and shipping disruptions exposed the other side of optimization.
Efficiency without resilience can create fragility.
Hormuz demonstrates it physically. Semiconductor shortages demonstrated it industrially. European dependence on Russian energy demonstrated it geopolitically. AI infrastructure may eventually demonstrate it digitally.
This does not mean every country should reproduce every capability.
It does mean infrastructure planners must evaluate a cost that traditional efficiency models sometimes underprice: the cost of losing access.
A second pipeline can look unnecessary until the first is damaged. Another supplier can appear expensive until the primary supplier stops delivering. Another export market can look less efficient until the dominant buyer begins using access as leverage.
Redundancy is not inefficiency when the alternative is vulnerability.
Resilience has a price.
Dependence also has a price.
Good strategy requires acknowledging both.
9. Space Is Part of the Modern Border:
Sovereignty increasingly extends upward.
Satellites support communications, weather forecasting, navigation, financial timing, agriculture, surveillance, missile warning and intelligence.
They are infrastructure.
During conflict, infrastructure becomes capability.
Reporting examined during this research described intensive Russian satellite activity over military and strategic sites in the Middle East. Ukrainian intelligence and supporting security sources argued that imagery was being shared with Iran. Russia did not confirm that claim, and satellite overflight alone does not establish that imagery was collected, transferred or used for a specific strike.
That distinction matters.
What is documented and what is inferred should never be allowed to blur simply because the stronger version is more dramatic.
The structural lesson does not require overstating the intelligence.
Orbital systems can influence events on the ground. Modern societies increasingly depend on them for civilian and military functions.
Satellites, launch capability, communications networks and Earth-observation systems therefore belong inside the sovereignty discussion.
The territory a country must protect no longer ends at its coastline.
Part of the infrastructure supporting that territory is orbiting hundreds of kilometres above it.
10. Money Has Infrastructure Too:
Payment systems usually become visible only when they fail or when access is restricted.
Every card transaction, international transfer, clearing process and financing arrangement travels through institutional infrastructure.
Banks, card networks, settlement systems, currency markets, correspondent-banking relationships, regulatory systems and sanctions architecture all form part of that machinery.
If access to those systems becomes conditional, financial infrastructure becomes leverage.
Carney explicitly included payment systems in his definition of modern strategic capability and proposed deeper Canada-Europe cooperation in financial services.
A country may possess factories, minerals and energy while remaining vulnerable if its companies cannot efficiently finance projects, insure shipments, settle transactions or access capital.
Finance is therefore not separate from infrastructure.
Finance is part of the infrastructure.
When economic instruments are increasingly used alongside traditional military and diplomatic power, financial networks become geopolitical terrain.
11. Europe’s Offer, Without Getting Ahead of the Facts:
On September 16, European Commission President Ursula von der Leyen proposed moving beyond the existing Canada-EU relationship toward what she called an “Alliance for the Future,” opening the possibility of Canada becoming the European Union’s first “associate member”.
The concept is unprecedented, and its precise legal and institutional meaning remains undefined.
Carney welcomed the ambition the following day.
That is what happened.
Several things have not happened.
Canada has not joined the European Union. Canada has not entered the EU single market. Canada has not joined Schengen. Canadians have not acquired automatic freedom-of-movement rights across Europe. There is no established treaty category of Canadian-style associate membership waiting to be activated.
By September 18, Reuters was reporting that European and Canadian officials still faced significant legal, political and practical questions in defining what such an arrangement could become.
Ground News’ comparison of coverage from across the political spectrum similarly shows uncertainty over the eventual meaning of the proposal.
That uncertainty should not be treated as weakness in the story.
It is part of the story.
The proposal is significant precisely because governments are attempting to design something that does not yet neatly fit an existing category.
12. The Substance Beneath the Headline:
The practical content of the emerging Canada-Europe relationship is more interesting than the label attached to it.
In Strasbourg, Carney proposed deeper cooperation across critical minerals, defence industrial capacity, AI and compute, energy security, space, payments, digital trade, financial services, research, education and connectivity.
He also discussed potential Canadian participation in Erasmus+ and the next generation of Horizon, cooperation on AI safety, pooled sovereign compute, critical-mineral processing, LNG and hydrogen infrastructure and deeper financial-market integration.
Viewed together, these proposals suggest a different way of thinking about alliance structures.
Instead of beginning only with geography, begin with capability.
Who has energy? Who has processing capacity? Who has capital? Who has compute? Who has research institutions? Who has defence manufacturing? Who has market scale? Who has the trusted legal environment?
Then connect the parts.
This does not automatically create sovereignty. It can create another form of dependence if poorly designed.
But diversified relationships among several partners can reduce the risk that any one partner acquires overwhelming leverage.
That is the strategic argument worth studying.
13. October Is a Checkpoint, Not a Finish Line:
The next EU-Canada summit is scheduled for October 29 and 30, 2026, in Montreal.
The summit matters.
But it should not be described as a ceremonial signing of a completed associate-membership arrangement.
The architecture is still being defined.
That means October should be treated as another checkpoint.
What has actually been proposed? What has been agreed? What requires legislation? What requires financing? What requires regulatory harmonization? What requires parliamentary approval? What requires construction? What has a delivery date? What remains aspirational?
This is why the Season 6 scoreboard matters so much.
ANNOUNCED means an intention has been made public.
AGREED means relevant parties have formally committed.
FINANCED means capital has actually been allocated.
UNDER CONSTRUCTION means implementation has physically or institutionally begun.
OPERATIONAL means the capability exists.
USED means the capability has demonstrated that it can perform the strategic function for which it was created.
A press conference lives near the top of that list.
Sovereignty lives much closer to the bottom.
14. CETA, Capital and the Infrastructure Already Beneath the Story:
Canada and Europe are not suddenly inventing an economic relationship because relations with Washington have become more difficult.
The underlying architecture has existed for years.
CETA has been provisionally applied since 2017.
In March 2026, Canada and the EU reported that two-way goods trade had increased by more than 75% since provisional application, while bilateral services trade had increased by approximately 97%; the two sides also formally launched negotiations toward a Canada-EU Digital Trade Agreement.
That changes how the current moment should be understood.
This is not a new house being built on empty land.
It is an existing structure being expanded because geopolitical conditions have changed.
The same distinction applies to capital.
Canada’s September Investment Summit brought together investors from nearly 30 countries managing more than C$100 trillion in assets and produced announcements representing nearly C$500 billion in new investment commitments, financing and capital mobilization across multiple time horizons, according to the federal government.
That figure is significant, but it is not C$500 billion already poured into operating infrastructure.
The government’s breakdown includes pension and institutional capital, long-term bank financing, investment vehicles and projects whose spending will unfold over years.
The real test comes later.
How much becomes actual investment? Which projects obtain permits? Which reach financial close? Which begin construction? Which eventually operate?
The government has also set an ambition to help catalyse more than C$1 trillion in total investment over five years through public, private and institutional capital.
That ambition should be measured by outcomes.
Sovereignty cannot be financed with announcements alone.
15. Diversification, Trust and Leverage:
Diversification is beginning to appear in Canada’s trade data.
Statistics Canada reported that in July 2026, merchandise exports to countries other than the United States rose 7.4% to a record C$25.6 billion, with non-US destinations accounting for 33.7% of Canadian merchandise exports that month.
That is meaningful.
It is also one month.
Commodity prices, exchange rates, economic cycles and trade composition all influence monthly data.
The responsible conclusion is therefore not that Canada has replaced the United States.
It is that alternative demand exists, and that diversification is becoming measurable rather than merely rhetorical.
Reuters has emphasized those limits as well, noting that the United States remains by far Canada’s largest market and that geography, industrial integration and shared security architecture make full economic decoupling unrealistic.
Diversification does not mean divorce.
It means portfolio management.
If one customer buys nearly everything you produce, that customer possesses significant leverage. If several markets become meaningful buyers, the bargaining relationship changes even if the original customer remains the largest.
Donald Trump’s September 17 response provides a concrete example of the mechanism. After von der Leyen proposed the new Canada-EU relationship, Trump described the idea as “laughable” and said the United States could impose heavy tariffs or restrict trade with Europe if he regarded the arrangement as hostile to American interests. No new tariff specifically tied to the proposal had been imposed at the time of writing.
The structural lesson does not require speculation about motive.
Market access can be used as leverage.
That is one reason diversification matters.
Trust enters the same calculation.
Long-lived infrastructure requires confidence that contracts will survive political transitions, courts will function, property rights will remain enforceable, regulations will remain intelligible and policy will not change without warning.
Investors therefore evaluate more than resources and tax rates.
They evaluate institutional behaviour.
A mine can operate for decades. A port can operate for generations. Data centres require billions of dollars before they return capital. Defence production depends on contracts that can outlive governments.
In that environment, predictability acquires economic value.
Predictability can become a form of national capital.
Countries can compete partly on whether other governments, investors, companies and citizens believe their systems will still function tomorrow.
16. Sovereignty, Managed Interdependence and Agency:
Globalization was built around efficiency.
Produce goods where they can be produced most effectively. Reduce barriers. Integrate supply chains. Allow specialization. Move capital. Connect markets.
That system generated enormous prosperity.
It also created concentrations of dependence whose strategic implications were often underpriced.
Natural gas can become leverage. Semiconductors can become leverage. Rare earths can become leverage. Food can become leverage. Shipping routes can become leverage. Payment infrastructure can become leverage. Cloud computing can become leverage. Market access can become leverage.
The answer does not have to be deglobalization.
A more useful concept may be managed interdependence.
Remain connected. Trade. Invest. Share research. Build alliances. Integrate markets where doing so creates mutual benefit.
But identify concentrations of dependence capable of becoming coercive and develop alternatives before a crisis forces the issue.
That is not isolation.
It is risk management at the level of the state.
It also reveals a paradox.
International cooperation can sometimes strengthen sovereignty rather than diminish it.
One supplier controls a strategic input and that supplier possesses leverage. Several suppliers provide the same input and the customer possesses options.
One market buys nearly everything a country exports and that market possesses leverage. Several substantial markets create bargaining room.
One country supplies an entire defence system and that country possesses leverage. Interoperable suppliers across several allied economies create alternatives.
Modern sovereignty may therefore depend less on eliminating dependence than on distributing it intelligently.
That brings us back to the larger structure of Season 6.
OPTIONS established that dependence creates leverage and alternatives redistribute it.
ALIGNMENT examined how countries connect those alternatives around shared interests.
ADAPTATION examined whether the alternatives can actually be financed, built and used.
SOVEREIGNTY asks what usable alternatives ultimately protect.
The answer is agency.
Not domination.
Not isolation.
Not the fantasy that a modern country can do everything alone.
The ability to make choices.
To cooperate because cooperation is beneficial rather than because there is no alternative.
To trade because trade creates prosperity rather than because one customer controls the economy.
To use technology without surrendering authority over critical systems.
To share infrastructure without allowing one provider to disable society.
To say yes because a country chooses to, and to retain the ability to say no without being economically or strategically incapacitated.
That is what the pipelines, ports, data centres, mines, satellites, payment networks, trade agreements and alliances ultimately protect.
They protect the space in which choice remains possible.
Perhaps sovereignty has never been as absolute as political maps made it appear. States have always needed neighbours, customers, lenders, allies and suppliers.
The twenty-first-century difference is the density of those connections and the speed at which dependence can travel through them.
A disruption in one strait can alter transportation costs on another continent. A vulnerability in a digital system can cross borders in seconds. A decision in one capital can reach a factory, farmer, investor or family thousands of kilometres away.
Healthy systems therefore require feedback before vulnerability becomes crisis. They require governments capable of recognizing concentration risk, institutions capable of changing course, markets capable of creating alternatives and alliances capable of sharing capacity without converting cooperation into domination.
Accountability in that sense is not revenge.
It is feedback.
It is the ability of a system to detect that something has become dangerously concentrated and respond before circumstances make correction far more expensive.
That is also why sovereignty should not be measured by how loudly a government declares independence.
It should be measured by how much meaningful choice remains when pressure arrives.
The test of sovereignty is not whether you need others. It is whether needing them prevents you from saying no.
Season 6 began with dependence.
It now arrives at agency.
Options create alternatives. Alignment connects them. Adaptation makes them usable. Sovereignty protects the right to choose among them.
And that right to choose may ultimately be the infrastructure beneath freedom itself.
Incorrigible Myths™ AI Disclosure:
The research, analysis, arguments, interpretations and conclusions published in Incorrigible Myths™ are my own.
AI is used only as an editorial aid for proofreading and refinement, including spelling, grammar, syntax, clarity, structure and readability. It is not used to originate the analysis, determine the argument, form the conclusions or replace my own research and judgment.
Because AI-assisted editing can influence linguistic patterns in a finished draft, Substack’s “Scan for AI” feature may occasionally flag portions of an article as AI-generated even when the underlying research, reasoning and authorship are my own.
For transparency, I retain original drafts, revised drafts and final publication copies as documentation of my authorship and editorial process. These materials can be made available upon reasonable request.
The views, interpretations and conclusions expressed in Incorrigible Myths™ are entirely my own and are presented for independent analysis, commentary and public-interest discussion. They do not represent the views of any employer, client, organization, institution or other entity with which I am currently or previously affiliated.
World Affairs Writer ✍🏼Geopolitics, AI, Humanitarian Risk AnalystWhistleblower • Incorrigible Myths™ 🇨🇦
This is not the time to know less. It is the time to understand more.
Incorrigible Myths is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Get full access to Incorrigible Myths at incorrigiblemyths.substack.com/subscribe
