AI stalls on the financial case long before it stalls on the technology. That is AnnMarie Castrigno's position after 23 years supporting the Intelligence Community and Air and Defense, most recently leading business area IT for the $7.8B+ Electronic Systems sector at BAE Systems, Inc. She spent those years translating IT spend into CFO aligned language, and her take on why AI investment keeps breaking down is direct.

CIOs are still using the wrong financial model, and they are the reason their own AI initiatives get killed.

"It's really hard to measure ROI for AI," she said. "You may not be saving money. You may be saving cycle time, or cost of poor quality, or contract deliverable time. It's not a one-for-one line in the sand."

Her replacement for ROI is investment decision confidence. In practical terms, it is the question a CFO answers every time an AI initiative comes up for more funding: fund the next round, or shut it down.

The CIO Owns the Translation

Castrigno is direct about where ownership sits.

"The CIO should own the value translation," she said. "The CIO sits in between IT and the business."

That ownership usually fails at the CIO level, not the CFO level. Most CIOs are still heavily grounded in technical expertise, and the role rewards depth. It also rewards a kind of financial literacy that does not come naturally with depth.

"Where the CIO is still more technically adept than business adept, that's where the breakdown is," she said. "The CIO has to bridge the relationship with the technical experts and the CTO, but the CIO also has to speak the language of business outcomes and CFO financial acumen."

A CFO's decision confidence runs on cost of IT per user, cost of IT as a percent of revenue, and cost of IT as a percent of operating expense. Most AI pilots report adoption saturation, code acceptance rates, and proof-of-concept benchmarks instead. Those reports are why a CFO kills the next round of funding, and the CIO is the one who handed over the case that made it easy.

Why She Has Credibility to Say This

Castrigno did not arrive at this position academically. She reduced IT spend from $86 million to $53 million in a defense-contractor environment where compliance and uptime were non-negotiable. That cut held.

The levers were specific. Expense consumables first: duplicate laptops, runaway mobile device costs, telecom circuits still billing on sites shut down years earlier. Duplicative software licenses next, consolidated down to one tool. Discretionary refresh cycles extended, labor reduced where possible.

The $33 million came out of line items IT had not made transparent to the finance side in the first place. That is the muscle behind her point about ROI. She has moved the numbers a CFO cares about, at scale, without breaking the business those numbers support.

The Cut That Looked Small and Wasn't

Her sharpest operational observation has nothing to do with AI.

In one enterprise IT organization, leadership eliminated a small team of ticket-queue monitors, a role that looked like pure overhead on a P&L. The monitors made sure incoming tickets routed to the right team on the first pass.

"These were people at low grade levels, low pay rate," she said. "But the impact to service was huge because of the time that was now lost in getting tickets to the right place."

The unit cost of a role tells you almost nothing about its operational leverage, and that is the lens every AI conversation needs.

The Role Most CIOs Are Already Absorbing

Castrigno describes herself in brand terms, which is unusual for a CIO. Early in her career, she took over a role delivering IT to a sector president who told her, "You know where the dead bodies are." Her edge was separating the emotion from the pain points and putting the fix on the table in that order.

That pattern compounded across 23 years. When a relationship was broken, she was the call. When governance was weak, she was the call. The role she has come to define is the translation layer between technical capability and financial accountability, and it is the role most CIOs today are absorbing whether they name it or not.

"Where enterprise IT is not a partner, it becomes nothing more than a cost center."