AI Is Scaling Faster Than Finance Controls 🤖

This week, the finance conversations that stood out were not about whether AI belongs in the function. That debate is mostly over. The harder question now is how CFOs evaluate very different AI use cases, how they keep cyber risk from turning into a financial event, and how they redesign finance work without confusing automation progress with operating maturity.

This issue covers the control challenge sitting underneath AI adoption, a practical way to distinguish between useful automation and unmanaged exposure, and three articles worth your time

As finance teams rethink where stability and upside can come from, it is also worth looking at how leaders are approaching diversification when the usual assumptions feel less dependable. Today’s partner, Masterworks, offers one more lens on that question before we get into this week’s numbers.

If you have enough money that you think about buckets for your capital…

Ever invest in something you know will have low returns—just for the sake of diversifying?

CDs… Bonds… REITs…  :(

Sure, these “boring” investments have some merits. But you probably overlooked one historically exclusive asset class:

It’s been famously leveraged by billionaires like Bezos and Gates, but just never been widely accessible until now.

It’s had attractive growth and hasn’t tended to move in line with other markets from 1995 to 2025.

It’s not private equity or real estate. Surprisingly, it’s postwar and contemporary art.

And since 2019, over 70,000 people have started investing in SHARES of artworks featuring legends like Banksy, Basquiat, and Picasso through a platform called Masterworks.

My subscribers can SKIP their waitlist and invest in blue-chip art.

Just click this link

^{*Past performance is not indicative of future returns. Important Reg A disclosures: masterworks.com/cd}

That number matters because it shows why cyber risk can no longer sit outside the CFO’s field of view. Microsoft argues that cybersecurity has become a financial leadership challenge as incidents increasingly translate into financial loss, operational disruption, regulatory pressure, and board-level scrutiny. For finance leaders, the real shift is that AI is not only changing productivity expectations. It is also accelerating the need to treat resilience, control, and incident readiness as part of capital discipline.

If cyber exposure can now affect revenue, operations, and investor confidence simultaneously, it is no longer just an IT issue. It is part of the finance risk model.

At one company, the AI discussion kept getting stuck because leadership was treating every initiative as if it belonged in the same bucket. A reporting automation tool, a forecasting model, and a more ambitious agentic workflow were all being discussed with the same language, the same approval logic, and the same vague promise of future efficiency. The problem was not a lack of enthusiasm. It was a lack of separation. Once the use cases were split by risk, cost profile, and control burden, the conversation finally became finance-grade instead of hype-grade.

When the CFO is asked which AI projects deserve more capital and which ones need tighter control, the answer comes from a structured portfolio view instead of one oversized AI narrative.

CFO.com argues that finance leaders should stop treating AI as one single ROI challenge and instead view it as a portfolio of different use cases with different costs, timelines, risks, and value profiles.

My take: This matters because finance gets sloppy the moment very different bets are forced into one approval model. A routine automation project and a transformational AI initiative should not be expected to prove value or carry risk in the same way.

CFO Dive reports that Microsoft now frames cybersecurity as a financial leadership challenge as cyber incidents increasingly create financial loss, operational disruption, and board-level pressure, with AI accelerating the stakes.

My take: Finance leaders should read this as a control story, not a security side note. Once cyber incidents can hit revenue, operational continuity, and investor confidence, the CFO cannot afford to treat resilience as someone else’s budget line.

At the AI for CFOs summit, Meta outlined how its finance organization used an 18-month process redesign effort and agentic AI to cut one invoice-editing workflow from fully manual handling to 7% manual intervention in seven days, while aiming to compress a 10-day procurement cycle into one day.

My take: The real lesson is not speed by itself. It is that the impressive result came after deep process redesign, clean data plumbing, and a clear view of where humans still belong, which is exactly the discipline many companies try to skip.

The most common place I see this break is when companies say they are investing in AI, but cannot show where each use case sits on the spectrum between task automation, decision support, and autonomous action. Once that happens, budget logic gets fuzzy, risk ownership gets blurry, and finance loses the ability to explain what exactly it has approved.

Can you produce, right now, a list of your finance AI use cases, what category each belongs to, what human review it requires, and what risk it introduces? If not, that ledger is your next 30-day project.

Percent's secondary marketplace lets accredited investors buy into eligible deals or indicate interest in selling existing positions. Secondary market access in private credit is still rare. 16.72% current weighted average coupon. Terms start at 3 months. New investors can receive up to $500 credit.

See how it works.

^{Alternative investments are speculative. Secondary liquidity not guaranteed. Past performance not indicative. Terms apply.}

The finance function is not short on AI ideas right now. It is short on clean ways to govern very different kinds of AI work without collapsing them into one vague strategy bucket. Gartner’s view on AI portfolios, Microsoft’s framing of cyber as a finance issue, and Meta’s process-first approach all point to the same conclusion: the control problem is becoming as important as the technology opportunity.

That is why this week’s pattern matters. One article shows that AI investments need to be judged as different bets. One makes clear that cyber risk now belongs inside the CFO’s field of responsibility. One shows that meaningful automation gains come after process redesign, not before.

The common thread is straightforward. Finance leadership in this phase is less about buying into AI and more about sorting, governing, and sequencing it properly. The teams that handle this best will not be the ones making the loudest claims about automation. They will be the ones who know which use cases deserve scale, which ones need tighter review, and which risks become more dangerous when speed improves faster than control.

Until next edition. — Marcus Reid

If your team has already built a clean way to classify AI use cases by value, risk, and control burden, reply directly to this email. I am collecting examples of the frameworks finance leaders are using to keep AI investment disciplined while the pressure to move keeps rising.