Planful's Analyst Assistant tackles Finance workflow friction - but the GENCFO Trends Report 2026 shows the real bottleneck is cultural

Finance teams are busier than ever, yet many struggle to translate that effort into strategic influence. A recent Planful webinar on Artificial Intelligence (AI) in finance, alongside findings from the GENCFO Trends Report 2026 (published in partnership with Planful), paint a consistent picture: the Finance function knows what it needs to do next, but getting there requires more than new tools.

Planful’s webinar, hosted by Cameron Moreau of the company’s customer marketing team, opens with its own survey data, drawn from over 450 global Finance professionals. 60% of Finance teams report already using AI daily, while 90% of leaders say they are on a path toward adoption. But Marina Ureche, Senior Product Marketing Manager at Planful, explains that adoption alone is not the same as impact. She warns:

What we’re seeing across teams is a growing gap between AI activity and AI impact. Simply using AI is not enough to be successful.

Finance teams have experimented widely with general-purpose AI tools like ChatGPT or Gemini, but generating an answer and trusting it enough to present to a board are different problems. As Ureche explains, citing Gartner research, AI only delivers value when it operates within governed Finance data with clear lineage and controls, not when it sits outside the planning system.

Purpose-built, not bolted on

The webinar defines four criteria for Finance-grade AI. Explainability: showing the dimensional drivers behind an answer, not just the output. Traceability: maintaining a clear audit trail so that any figure can be walked back through the consolidation path to its source transaction. Augmentation: surfacing insights for a human to validate, rather than automating judgment out of the loop. Last but not least, governance: operating within the same role-based permissions, approval workflows, and encryption standards that Finance already relies on.

Planful’s Analyst Assistant, now generally available, is the company’s answer to these criteria. Architecturally, it pairs OpenAI’s natural language processing with Planful’s own patented calculation engine – meaning the Large Language Model (LLM) handles query interpretation and narrative generation, while the financial computation runs through Planful’s existing engine with its fiscal calendar logic, chart of accounts structure, and dimensional security. When a user asks “What were the drivers of operating expense variance to plan last quarter?”, the response includes not just a number but a narrative breakdown by cost center, department, or account – scoped to only the dimensions that user has permission to see. The data stays within the platform, end-to-end encrypted, and is never used to train external models.

Matt Sledge, Senior Product Marketing Manager at Planful, underlines what separates this from a generic chatbot: 

Every response includes explanations, the driver-level detail that shows not just what changed, but why it changed.

Sean March, from Planful’s solution consulting team, demonstrates this during the webinar. The Analyst Assistant surfaces variance drivers across departments, flags cost centers exceeding budget thresholds by a configurable margin, and generates trend visualizations from conversational prompts. He shows that when the assistant lacks sufficient data to answer a question, it asks a clarifying question rather than hallucinating a response – an important design choice for any AI tool but especially for a Finance context where a confident wrong answer is worse than no answer at all.

A Planner Assistant is next on the roadmap, designed to seed forecasts and run what-if scenarios directly from the same governed data environment. March also demonstrates predictive forecasting capabilities already available in beta: the system trains on historical actuals, applies one of seven selectable statistical models, and generates forward-looking projections that can be pushed directly into budget templates. A bring-your-own-algorithm capability is planned for later in the year. The broader roadmap includes further persona-based assistants for controllers, modellers, and administrators – each scoped to the workflows and permissions of that role.

Where the research aligns

The GENCFO Trends Report 2026, produced in partnership with Planful and based on polling of 150–200 mid-to-senior Finance professionals, adds an extra dimension to the webinar’s product narrative. Its headline finding is a 23-point perception–performance gap. 44% of respondents describe their Finance function as a “strategic business partner,” but when asked to rate the behaviors that support that claim, only 17% actually demonstrate it. Over half remain desk-based, producing reports with minimal business collaboration.

GENCFO founder Christopher Argent describes this as a Dunning-Kruger dynamic specific to Finance – teams believe they are performing well because they lack visibility of what “great” looks like. The report’s other findings reinforce the pattern. Technology overtakes macro-economic volatility as the number one external driver (41%), yet 66% of organizations are still only exploring or experimenting with AI. Just 13% are implementing or scaling.

The burnout finding is helps to emphasize who this is affecting the most. 49% of respondents identify burnout and disengagement as their top human capital concern – well ahead of resistance to digital adoption (25%) or talent gaps (12%). The report concludes that transformation layered on top of demanding business-as-usual workloads is exhausting teams that have not re-designed their ways of working.

Tools solve for friction, culture solves for impact

Taken together, the webinar and the report tell a complementary story. Planful's product narrative explains that governed, embedded AI can remove operational drag from finance workflows – fewer manual steps, faster access to trusted data, clearer explanations, and role-aware insights. The GENCFO research validates the realities on the ground – 71% of respondents say poor systems and processes are the biggest barrier to automation, and 59% cite data and systems as the primary blocker. On those terms, the Analyst Assistant addresses a genuine pain point.

Reporting volume has increased without a corresponding increase in decision quality. Finance teams are producing more outputs faster, yet the business still experiences Finance as operationally helpful rather than strategically influential. As the report puts it: automation without re-design simply funds more reporting, not more impact. When asked which non-technical skills will be most critical in 2026, respondents split almost evenly across communication (39%), cross-departmental collaboration (29%), and data storytelling (26%) – none of which are solved by faster variance analysis alone.

That tension runs through the entire AI-in-finance conversation right now. Vendors can and should build tools that eliminate manual drudgery – Planful’s demo was credible on that front. But unless the freed capacity is deliberately reinvested in strategic engagement, efficiency gains compound the existing model rather than changing it.

My take

Planful is not overselling AI as transformational on its own. The emphasis on explainability, governance, and augmentation is precisely calibrated for a Finance audience that has seen too many tools launched with a “here it is, use it” attitude (to borrow a phrase from the GENCFO report’s own interviewees). The Analyst Assistant’s architecture – grounded in the customer’s own data, respecting existing security controls, producing auditable outputs – directly addresses the trust deficit that 93% of GENCFO respondents flagged: limited or no confidence in their organization’s AI governance.

But what happens after the time is freed? The report findings plainly state that the perception–performance gap in Finance is behavioral and organizational before it is technical. If business partnering remains a job title rather than a way of working, if reporting volumes are not deliberately cut, and if automation simply accelerates existing outputs, then even the best AI assistant will not close the gap.