AI tools like Copilot, Gemini, and ChatGPT are marketed as assistants, time savers, and productivity boosters. In reality, they are new gateways to knowledge. Instead of clicking links across the open web, you get a single synthesized answer. Instead of navigating multiple SaaS tools, your AI sits between you and your data.
This shift centralizes power. Whoever controls the AI interface decides what you see, what’s excluded, and how your decisions are nudged.
AI as Bait
Think about Copilot, Gemini, or ChatGPT. They’re marketed as “assistants,” but their strategic role is bigger: they’re the front door. These tools draw users into ecosystems where the true monetization lies: cloud infrastructure, enterprise licensing, and data services.
- Microsoft doesn’t expect every consumer to pay $30 per month for Copilot. What it does expect is for enterprises to bundle Copilot into Office 365, locking entire workforces into the Microsoft stack for years to come.
- Google uses Gemini not just to enhance Gmail and Docs, but to keep users tied to Workspace while redirecting ad-spend into AI-augmented Search.
- Oracle infuses AI into databases and ERP systems to protect its lucrative enterprise workloads from defecting to rivals.
AI isn’t the product; it’s the sugar coating that ensures you keep feeding the machine your work, your clicks, and your habits.
The True Currency: Data Centralization
For decades, the internet has been (relatively) open. You searched on Google, clicked to a website, and consumed information at the source.
Generative AI flips that script. Now the AI itself provides the answer. Every time you use these systems, two things happen:
- They harvest data from you. Your queries, files, and behaviors become training fodder.
- They shape what you see. Instead of navigating an open web, you consume through a centralized filter.
Over time, this creates what economists call a chokepoint market. The AI interface becomes the gatekeeper, controlling not only what information flows through but also what gets left out. It’s a massive consolidation of power, disguised as convenience.
From Search to Gatekeeping
Search engines already concentrated information power, but at least they still pointed outward.
Generative AI goes further: it removes the need to click at all.
- When you ask Google Gemini for an answer, it may summarize results rather than sending you to 10 different websites. Publishers lose traffic; Google gains control.
- When you ask Microsoft Copilot about a project, it synthesizes content across Jira, Teams, and Dropbox. Third-party SaaS tools become invisible.
- When Oracle infuses AI into finance and HR systems, it ensures decision-makers stay inside Oracle’s suite rather than considering external analytics vendors.
The outcome: instead of 100 sources of truth, you get one centralized intermediary. That intermediary belongs to a handful of mega-corporations.
The Cloud Angle
If the AI front-end is bait, the real monetization engine is cloud compute.
Training and running AI models consumes staggering resources — GPUs, storage, networking. Microsoft Azure, Google Cloud, and Oracle Cloud aren’t investing in AI altruistically. They’re ensuring that every enterprise experiment with AI drives more spend into their cloud platforms.
AI, in this sense, is the perfect growth engine for cloud vendors:
- Customers generate more data.
- AI workloads require more computation.
- Regulatory and security complexity make switching providers nearly impossible.
The bigger the AI hype, the deeper the enterprise sinks into one vendor’s infrastructure.
The End Game: Ultra-Centralization
Strip away the marketing gloss, and the end game looks like this:
- Information Chokepoints
Instead of navigating a diverse web, consumers interact with a handful of AI systems. Those systems decide what’s worth surfacing — and what’s not. - Cloud Lock-In
Enterprises pay escalating fees to run AI workloads on a small number of hyperscaler clouds. Exit options dwindle. - Behavioral Shaping
If the AI assistant recommends one product, vendor, or policy path over another, it subtly nudges human decision-making. At scale, this centralizes influence in ways traditional advertising never could. - Regulatory Insulation
With control of platforms and data, Big Tech can shape regulatory frameworks to favor their architectures, pushing smaller competitors out of the market.
The outcome is ultra-centralization of knowledge, compute, and decision-making — with AI as the friendly user interface that makes it all palatable.
Why This Matters for Consumers
For individual users, the trade-offs are subtle at first:
- Benefits: smoother apps, fewer clicks, faster answers.
- Costs: loss of diversity in information sources, creeping subscription fees, and constant data harvesting.
Over time, the risk is that consumers lose autonomy. If you never leave the AI interface, you stop seeing alternatives. Choice erodes not by censorship, but by friction — if the answer is right in front of you, why go searching further?
Why This Matters for CIOs and Enterprises
CIOs should be even more cautious. AI hype often masks vendor dependency risk.
- License inflation: Copilot-style add-ons raise per-user costs, often without proven ROI.
- Data residency & sovereignty: Centralizing AI in one vendor’s cloud can create compliance nightmares.
- Strategic blindness: If AI intermediaries surface insights only from within their own stack, enterprises risk narrowing their view of reality.
The CIO’s mandate should be clear: adopt AI where it drives measurable outcomes, but avoid ceding control of your entire data estate to a single gatekeeper.
Conclusion: The Illusion of Progress
AI is real technology with real potential. But the money flowing into it isn’t a charity project to make our lives easier. It’s a power play.
The real prize isn’t smarter machines — it’s centralized control of data and the flow of information.
If we’re not careful, we’ll wake up in a world where instead of consuming knowledge from a rich ecosystem of sources, everything passes through one of three or four corporate filters. And at that point, AI won’t be serving us; it will be serving its owners.