Finance’s AI moment has arrived. Two-thirds of finance professionals say they’re using or piloting AI. Their confidence is trending upward, and the appetite for deeper integration is real. But according to new data from the Yooz 2026 AI in Finance Report, only 10% of finance teams say AI is embedded in their core processes.
The data reveals a failure of operationalization, which points to the next challenge facing finance organizations. Adoption is already underway, but the next step is building the infrastructure and process foundations that turn AI experimentation into reliable, embedded capability.
Embedded, not Experimental
The distinction between using AI and embedding it is more than semantic. A team that runs a reporting query through an AI tool is using AI. A team whose invoice approval workflow automatically screens every transaction for duplicate entries, vendor anomalies, and out-of-policy amounts before any human reviews it has embedded AI. The first produces an output. The second changes how the organization controls its financial operations.
Most finance teams are still in the first category. The survey finds that 37% say AI is used in a few specific areas, and another 20% are piloting or testing it. That accounts for the bulk of the 67% using or piloting figure. Both are valid stages of the journey, but neither represents the standardized, repeatable workflow integration that delivers consistent outcomes at scale.
Getting there requires organizations to do the foundational work that makes AI effective. That work is primarily about process, not technology.
Process First, then AI
One of the clearest lessons from organizations that have successfully moved from pilots to embedded AI is that the technology performs best when the underlying workflows are already well-defined. AI amplifies what’s there. If approval chains are inconsistent and GL coding varies by individual, AI will inherit that inconsistency and scale it.
The right sequence is to map how work actually flows today, identify where the standard path breaks down, and establish clear rules for how exceptions are handled before layering AI on top. This is the operating logic behind Lean Financial OperationsTM, the framework Yooz uses to describe the model finance teams should be building toward. It centers on standardized processes, embedded automation, and controls that run continuously rather than being applied manually at review points.
For AP specifically, that means defining the touchless path for routine invoices, establishing clear thresholds for human review, and building fraud detection into the approval workflow rather than treating it as a separate function. When AI is integrated into a well-designed process, it doesn’t add complexity. It removes friction.
Enablement Is the Missing Piece
The survey data is unambiguous about what’s actually slowing progress. Lack of training or education tops the barrier list at 26 %, followed by lack of trust in AI outputs at 25%. Budget constraints land at just 10% and regulatory concerns come in at 12%.
Finance teams aren’t being held back by external forces. Addressing gaps in training and AI infrastructure is within their control. It takes an intentional approach to AI literacy. Teams need to understand where AI is already operating in their systems, what it’s doing, and how to interpret its outputs. That kind of foundational understanding is what converts cautious curiosity — the dominant mindset among 42% of respondents — into confident, consistent daily use.
It also requires guardrails that match the risk profile of each workflow. Flagging a duplicate invoice is a different stakes proposition than approving a six-figure vendor payment. The controls, review paths, and escalation rules should reflect that distinction. When teams can see that AI is operating within clear, well-defined boundaries, trust builds faster.
Leadership Has To Own It
The data points to a structural issue in who is leading AI adoption. Twenty-four percent of respondents say that IT or technology teams are the primary driver of AI adoption in finance. Only 13% point to the CFO or VP of Finance. Another 22% say no one in particular is leading the effort.
At the same time, the opportunity for finance leaders is significant. Interest in AI is growing, and confidence is starting to build, creating a clear opening for finance to take a more active role. Without that leadership, teams may have access to AI tools, but those tools often remain disconnected from core workflows and lack the training and governance needed to scale effectively.
Finance leaders bring a perspective that is well suited to this moment. They understand how work flows across the organization and how outcomes tie back to business performance.
When finance leaders take ownership of building the process foundations that support AI, it moves from isolated experimentation into something embedded in day-to-day operations.
The macro analyst desk brings highly sought after financial news based on market analysis, insider news and company filings.