AI as a Backdoor Price Hike
The dispute over Microsoft 365 price hikes and Copilot subscription bundling is a conflict about whether software makers can quietly bolt AI features onto existing plans, raise fees, and treat silence or confusion as customer consent. Italy’s competition authority, AGCM, has opened a regulatory investigation into how Microsoft informed customers about changes to its Microsoft 365 subscription pricing, after AI services were folded into the product. At the core of the case is a blunt question: did Microsoft give users a fair chance to say no, or did it design a system where most people would pay more without realising what they had signed up for? That question matters far beyond one jurisdiction or one product.
AGCM says Microsoft may have delivered information about the Microsoft 365 price hike in a “fragmented manner”, without making it “sufficiently clear” that Copilot and Designer had been integrated into the service. In other words, AI was not an optional extra; it was woven into a more expensive bundle and presented as the new normal. According to AGCM, subscribers were automatically moved onto a costlier plan and had to explicitly withdraw if they did not want to pay more. That is the opposite of informed consent. It treats inaction as approval, even when the terms of the deal have materially changed. If this becomes the template for AI feature pricing transparency, users will lose control over their own software budgets.
Defaulting Users Into Copilot Is Not Choice
The most troubling allegation is not that Microsoft added Copilot, but that it used default settings to nudge users into a more expensive tier. AGCM reports that consumers were placed on a new, costlier Microsoft 365 plan by default unless they explicitly opted out. Another description from the authority says subscribers were “automatically shifted” onto the more expensive option, leaving them to opt out if they did not want to pay extra. That is a classic dark pattern: make the pricey path effortless and hide the exit in fine print. When a core work tool changes terms in this way, staying put becomes a financial trap.
AGCM goes further and suggests this conduct “may also constitute an aggressive practice” because it appears to have “unduly restricted consumers’ freedom of choice”. That is a strong phrase, and it should be. Choice is meaningless if the default is secretly costly and the information is scattered across emails, dashboards and policy updates. If the regulator’s view holds, this is not a minor communication slip; it is a systematic attempt to convert user inertia into revenue. Microsoft, for its part, says it is committed to complying with local consumer law and will cooperate with the preliminary investigation. Cooperation is welcome, but it does not erase the fact that many users only “chose” Copilot because they were moved first and asked questions later.
The Real Issue: AI Feature Pricing Transparency
This regulatory investigation pricing case is not an attack on AI itself. AGCM has been explicit that its concern is not the inclusion of AI in Microsoft 365, but whether customers were clearly told that Copilot and Designer had arrived—and that their subscription bills would rise as a result. According to the authority, the changes were communicated in a fragmented way that failed to spell out what subscribers were getting for the higher price. That matters because AI feature pricing transparency should be straightforward: if you want to charge more, you say so, in plain language, before the new bill lands. Anything less looks like an attempt to hide a Microsoft 365 price hike behind glossy product marketing.
In AGCM’s view, Microsoft failed to give consumers enough information to assess what had changed and decide whether to renew their subscriptions. When the information gap is that wide, user consent becomes a legal fiction. At the same time, this case lands as Microsoft has been weaving Copilot into almost everything it sells, with pricing close behind. Other authorities are also scrutinising how Microsoft bundles and licenses its software ecosystem. Put bluntly, regulators are drawing a line: adding AI is fine; sneaking it in with confusing communication and forced upgrades is not. If software vendors want trust, they need to stop hiding new price tags behind “smart” features.
A Pattern of Bundling—and a Warning to the Industry
This is not the first time Microsoft’s product bundling has triggered scrutiny. A prior investigation into its suite led the company to agree to unbundle certain services to address competition concerns about cloud-based communication and collaboration tools. More recently, another authority signalled support for designating Microsoft’s Azure as a gatekeeper under the Digital Markets Act, a law that imposes obligations on dominant digital platforms. Now AGCM’s focus shifts from market power to the fine-grained reality of how Microsoft treats individual subscribers. The message is clear: regulators are no longer satisfied with high-level promises about choice; they want to see it in the click paths and billing flows.
For ordinary users, this case is a reminder that AI is often sold as magic, but billed as a mandatory upgrade. For the wider software industry, it is a warning. If you weave AI into every product and treat the rollout as a chance to ratchet up subscription revenue, you must disclose the cost clearly and leave the user in control. Anything else risks being labelled an aggressive practice that restricts freedom of choice. The safest long-term strategy is boring transparency: clear notices, genuine opt-ins and pricing that stays visible even when the marketing buzz of “Copilot everywhere” fades.
