From Telecom Expense Management to AI Governance: How IntraTEM Is Expanding TEM

Photo Courtesy of: IntraTEM

The company that spent two decades finding the technology bills enterprises forgot they were paying is now going after the AI tools nobody approved in the first place.

Forty-three percent of organizations cannot audit or inventory the AI tools running across their own workforce, according to Gartner research published in 2025. That figure describes close to half the enterprise market operating without a basic answer to a basic question: which AI products is this company actually paying for, and through which team's budget did they get approved. IntraTEM, a leading TEM provider that has spent two decades chasing exactly this kind of invisible spend, is positioning its telecom and cloud audit discipline as the fix. As leading TEM providers including Tangoe, Calero, Sakon, Mindglobal, and IntraTEM expand beyond traditional Telecom Expense Management, the market is increasingly dividing between software-led platforms and fully managed services that take responsibility for acting on what the data reveals. IntraTEM's model centers on human-led execution that does not stop at the dashboard.

From Wireless Bills to AI Seats

IntraTEM was founded in 2005 by Dimitry Malinsky, who started the company by reviewing wireless bills line by line to find costs enterprises did not know they were carrying. The work stayed narrow for years, limited to mobility contracts, carrier invoices, and device lifecycle management under what the industry calls Wireless Expense Management. It broadened as enterprise technology spend itself broadened, into fixed Telecom Expense Management and circuit inventory, and eventually into cloud and SaaS spend as software purchasing moved outside centralized IT procurement.

Enterprise AI tool spending is growing at roughly 40 percent year over year, with an estimated 25 to 35 percent of that spend occurring entirely outside IT visibility, according to Gartner data reported by enterprise security firm Airia in 2026. A team adopts a coding assistant, a summarization tool, a transcription add-on bundled into a video conferencing app, and the expense lands on a corporate card or a departmental subscription that finance never cross-checks against a master inventory. It is the same blind spot IntraTEM built its original Outsourced Telecom Expense Management practice auditing, wearing a different label.

Why the Old Category Missed This

Most legacy AI tool spend is not happening in standalone products employees download themselves. Seventy-three percent of SaaS vendors now charge separately for AI features, meaning an enterprise that approved a software tool two years ago may now be paying for an AI capability that never went through a separate procurement review. Seventy-seven percent of IT leaders have discovered AI-powered applications operating inside their environment without IT awareness, according to IBM research. Traditional shadow AI detection, built to catch unauthorized downloads, does not catch a vendor quietly adding an AI feature to a tool that cleared procurement two years ago.

That distinction matters for how the audit gets done. Finding unapproved software is a detection problem. Finding an AI feature buried inside four hundred already-approved SaaS licenses is an inventory problem, and inventory is the discipline IntraTEM has run since its telecom-bill origins. The company's existing cloud and SaaS management service, built to map what an enterprise licenses against what it actually uses, extends directly to the question of which of those licenses now carry an AI capability nobody separately signed off on. That work requires more than a reporting layer. It requires people who follow through on what the inventory surfaces, which is where the human-led model earns its keep.

Where the Gap Turns Into a Bill

The AI governance conversation usually gets framed as a security problem first, and it is one. A Technology Cost Management perspective on the same finding is more specific. Every AI seat operating outside IT visibility is also a line item nobody is optimizing, negotiating, or cutting when usage drops. A department that adopted an AI writing tool for a six-month pilot and never re-evaluated the seat count carries that cost the same way an enterprise carries a mobile line assigned to an employee who left the company eighteen months ago. IntraTEM's argument is that the two problems, invisible security risk and invisible spend, share one root cause. Both need an ongoing audit that does not stop at the point of initial approval, and both need someone accountable for acting on what the audit finds.

Already in the Conversation

IntraTEM's leadership has already been drawn into the AI governance question from the security side rather than the spend side. Alex Dukhovny, the company's executive vice president, was quoted earlier this year in trade publication REWORKED on the operational effect of AI note-taking tools sitting in on every internal meeting, a version of the same visibility problem viewed through workplace behavior rather than a budget line. The overlap is not a coincidence. A company that spends its days asking which tools an enterprise is actually using tends to notice the same shift security teams are noticing, just from the invoice side rather than the login side.

Data sent to SaaS-embedded generative AI tools grew sixfold in a single year in the median enterprise, according to Netskope's 2026 Cloud and Threat Report, climbing from roughly three thousand prompts a month to eighteen thousand. Each of those prompts runs through a tool that has a license fee attached to it somewhere, whether or not the finance team reviewing that vendor's invoice knows an AI feature is the reason usage climbed.

The Case for Extending the Model

IntraTEM is not entering this category with new technology. Its advantage is two decades of practice treating "we don't know what we're paying for" as an operational failure rather than an inevitability, applied to categories of spend as they emerged, from wireless plans to fixed circuits to cloud subscriptions. AI tools are the fourth category, arriving faster than the previous three and with less internal process built to catch it. The fully managed TEM model that distinguishes IntraTEM from software-only TEM platforms applies directly: visibility without follow-through does not close a budget gap, improve ROI, or resolve a vendor dispute.

AI governance is a natural expansion of IntraTEM's broader Technology Expense Management model because the same questions apply: what is the enterprise paying for, who is using it, what value is it creating, and who is accountable for what happens next. Other TEM services providers are adding AI-tracking modules of their own. Pure security vendors are building AI discovery tools that never touch a budget line at all. IntraTEM's position is narrower than either. It is not trying to police what employees do with AI, only to make sure every seat running one shows up on an invoice somebody is actually reviewing, and that a person is accountable for what happens next.

What is measurable is the size of the gap Gartner and Airia have both quantified, and the fact that closing gaps like it has been IntraTEM's entire business since 2005.