The AI Entrapment Strategy
How the subscription economy playbook became government policy.
China just mandated AI education for every child starting at age six. The UK committed £2 billion to AI infrastructure this year alone. Europe launched a €200 billion investment plan. Meanwhile, the Pentagon signed a $15 billion AI contract that runs for the next decade.
That's a lot of money flowing toward technology that most experts agree is currently overvalued. OpenAI burns through $5 billion annually while carrying a $100 billion valuation. Nvidia trades at 40 times sales. Industry veterans are calling it worse than the dot-com bubble.
So why are our governments rushing to lock in long-term contracts during peak bubble pricing?
When Adobe Creative Suite Became Rent
Adobe figured something out in 2013 that changed everything. Instead of selling Creative Sutie for $1,299 upfront, they started renting it for $52 per month under Creative Cloud. Customers hated the change initially. Online forums exploded with complaints about losing ownership of software they'd paid for.
But Adobe understood the maths. Someone using Creative Cloud for three years would now pay $1,872 instead of $1,299. Over ten years? That's $6,240 versus a one-time purchase.
The subscription model did something else crucial, it made switching costs enormous. Your files lived in Adobe's cloud. Your workflow integrated with their other apps. And your team trained on their interface. Leaving meant rebuilding everything from scratch.
Revenue jumped from $4 billion in 2013 to $19 billion by 2023. Adobe had transformed from selling software to renting access. Once customers were locked in, prices could increase gradually without triggering mass exodus.
The Government Spending Acceleration
Something similar is happening with AI contracts, except the customers are taxpayers who can't cancel their subscriptions.
The numbers tell the story. In just two years since ChatGPT launched, the Department of Defense awarded $670 million in AI contracts to 323 companies. The Department of Homeland Security added another $22 million across 20 firms. That represents a 20% increase from the previous two years.
The Pentagon's new "Advancing AI Multiple Award Contract" framework is worth $15 billion over the next decade. Individual companies like xAI, Google, OpenAI, and Anthropic each received contracts worth up to $200 million.
Similar patterns emerge globally. The UK allocated £2.4 billion in AI contracts between 2018 and 2024, with nearly half going to a single Microsoft deal worth £1 billion. The European Union launched AI factories backed by €2.1 billion across seven consortia. Their broader InvestAI initiative commits €200 billion through 2025.
These aren't pilot programs or research grants. They're decade long operational commitments signed during what multiple economic analyses call unsustainable valuations.
Legacy Lessons We Should Remember
Government technology contracts have a pattern. Systems designed to be temporary become permanent. What starts as innovative becomes obsolete but irreplaceable.
The IRS still runs 60 million lines of COBOL code dating to 1959. Social Security Administration uses similar programming languages that computer science departments stopped teaching decades ago. Federal agencies collectively spend $337 million annually maintaining legacy systems aged 8 to 51 years.
A Government Accountability Office study found agencies spend $15 billion - 30% of their annual IT budget - on non-competitive contract renewals. Companies like Microsoft and Oracle receive 25-30% of their government sales through these automatic renewals.
Once systems integrate deeply enough, replacement costs exceed maintenance fees indefinitely. Vendors understand this dynamic.
The Web Platform Warning Signs
We've also seen this playbook before in the private sector. Wix and Squarespace built empires on vendor lock-in that would make Adobe proud.
Both platforms make data export nearly impossible. Moving a website built on either service to another provider requires rebuilding from scratch. Millions of businesses stay trapped not because the service excels, but because switching costs are prohibitive.
The platforms know this. Their business model depends on customers' inability to leave. Initial pricing attracts users, then gradual increases extract maximum revenue from captive audiences.
WordPress offers an interesting contrast. Because it's open-source, businesses can migrate between hosting providers freely. Competition keeps prices reasonable and service quality high. No vendor can hold customer data hostage.
Government AI contracts aren't following the WordPress model.
Bubble Timing Considerations
What if the current AI bubble creates perfect conditions for vendor lock-in rather than preventing it? Companies burning billions annually need revenue streams that outlast the hype cycle. OpenAI's $5 billion yearly losses require sustainable income sources. Government contracts provide exactly that - guaranteed payments regardless of market conditions.
The timing aligns suspiciously well. Nvidia's price-to-sales ratio hit 40 times revenue. Palantir trades at 69 times sales. These valuations mirror dot-com bubble peaks before the 2000 crash.
Yet our governments sign decade long contracts at precisely these inflated prices. When the bubble bursts and private funding disappears, public sector commitments become companies' primary revenue source.
Adobe's subscription transition offers a template. Make the initial switch during customer resistance, then capture higher lifetime value once they're locked in. Government agencies will have even less flexibility than Adobe users when switching becomes necessary.
International Competition Pressure
China's mandatory AI education creates an urgency that bypasses normal cost benefit analysis. No democratic government wants to appear "behind" in technological competition. The national security framing makes questioning AI spending politically difficult.
European Union officials explicitly mention competing with China and the United States in their AI investment announcements. The UK government positions its spending as essential for maintaining relevance against bigger economies.
This competitive pressure removes traditional procurement safeguards. When national prestige is at stake, asking whether decade-long contracts at bubble prices make economic sense becomes unpatriotic.
The Urgency Manufacturing Machine
The competitive pressure gets amplified by something more calculated: manufactured urgency. AI evangelists like Sam Altman warn that artificial intelligence will replace entire industries within years, creating societal transformation on an unprecedented scale. Yet when you examine what companies actually market, the story changes completely.
Microsoft sells AI productivity tools that help write emails and summarize documents. Google markets improved search and ad targeting. No business advertises that AI will replace their entire workforce - that would be commercial suicide. The gap between public warnings about AI's world changing potential and private marketing of narrow automation tools reveals something important about how hype becomes policy.
This doom-mongering serves multiple purposes beyond generating headlines. It creates fear of missing out that encourages uncritical adoption. If the narrative suggests AI will fundamentally alter society within years, questioning decade long contracts becomes politically risky. Who wants to be the official who "held back" their country's AI capabilities?
The psychological mechanism mirrors Adobe's subscription transition perfectly. Create urgency around change, make the current model seem obsolete, then lock customers into long-term commitments before they fully understand the implications. Government procurement officers face the same pressure Photoshop users felt in 2013—adapt now or risk being left behind.
The contradiction becomes obvious when you step back: if AI truly poses the existential risks these evangelists describe, why accelerate deployment into every government system simultaneously? Unless the real goal isn't managing risk but capturing market share during peak receptivity to change.
When Maintenance Becomes the Product
Once AI systems integrate into essential government functions, vendors shift from selling innovation to selling continuity. The most profitable customers aren't those seeking cutting edge features - they're those who can't afford system failures.
Legacy IT contractors understand this dynamic perfectly. Companies maintaining decades old government systems often operate more profitably than those developing new solutions. Innovation requires investment and carries risk. Maintenance generates predictable revenue with minimal expense.
AI companies currently burning cash on research and development will eventually pivot toward this model. Government contracts signed during the bubble provide the customer base for that transition.
The switching costs built into current AI contracts - proprietary data formats, custom integrations, specialized training - mirror the vendor lock-in strategies Adobe perfected. Except governments can't simply choose different software when the prices increase.
Full Circle Economics
Maybe we're looking at this backwards. What if the bubble creates perfect conditions for vendor lock-in rather than preventing it?
Adobe's subscription success came from transforming customers who paid once into customers who pay forever. The initial resistance to change became irrelevant once switching costs made alternatives impractical.
Government AI contracts follow identical logic. Agencies that might have purchased software licenses now rent access to cloud-based systems. Integration requirements make future competition nearly impossible. Long-term agreements lock in bubble pricing for years.
When private AI funding eventually dries up, government contracts become these companies' primary revenue source. Maintenance fees replace innovation spending. What started as cutting-edge technology becomes another legacy system too expensive to replace.
The pattern suggests we're not witnessing a technology revolution disrupted by a financial bubble. We're watching the subscription economy scale to government budgets during optimal market conditions.
Whether that's a brilliant business strategy or massive public waste depends entirely on who's writing the checks.
The pace at which all of this is happening is certainly concerning, and it's going to change humanity in ways we can't even imagine.
This was a great read 💯 and very insightful too.