Professionals are backing away from AI for financial models

A July 2026 report from the Infrastructure and Project Finance Association says AI in financial modeling is still being held back by governance, auditability and verification gaps. That is helping drive demand for downloadable, professionally built templates that are easier to review, defend and use in regulated settings.

Why it matters: - Financial models are used to support funding, valuations and board decisions. - In regulated industries, outputs must be traceable, auditable and easy to defend. - The gap between AI capability and deployability is pushing professionals toward templates with visible formulas and documented assumptions. - That shift affects banks, investors, startups and other teams that need reliable models now, not experimental tools.

What happened: - A July 2026 report from the Infrastructure and Project Finance Association’s Special Interest Group on Artificial Intelligence in Financial Modelling, chaired by Forvis Mazars, found that AI adoption in financial modeling is being slowed by assurance concerns. - The report drew on roundtables with 57 participants from banks, rating agencies and government agencies in Europe, the Middle East and Africa and Asia-Pacific. - A May 2025 analysis in Financial Management Magazine by Liam Bastick tested tools including ChatGPT and Microsoft Copilot on real financial modeling tasks. - eFinancialModels said professionals are turning to downloadable financial model templates because AI tools still fall short on transparency and control.

The details: - The Forvis Mazars report says a more capable AI model is not automatically a more deployable tool. - The report says practical value depends on integration, governance and control as much as raw performance. - The research concludes that meaningful embedding of AI into live financial modeling workflows remains limited. - Bastick’s analysis found AI-generated models often create hard-coded outputs instead of dynamic formulas. - Users had to rebuild models manually to verify calculations, which reduced the practical value of the tools. - Bastick said that rebuilding a model to check it defeats the purpose of automation. - The testing also found illogical period treatment, inconsistent calculations and unnecessary repetition across periods. - AI tools struggled with one-off calculations such as terminal values in discounted cash flow analysis and final loan repayments. - The tools often repeated computations across all periods even when business logic did not require it. - The Forvis Mazars report flags “shadow artificial intelligence” risk when staff use consumer AI tools and reintroduce the outputs into company models without an audit trail. - That creates no traceability for how the output was produced or which assumptions were used. - Participants said that approach is not acceptable for banks and regulated institutions. - Industry-specific template platforms offer visible formulas, auditable assumptions and documentation that supports stakeholder review. - eFinancialModels said professionals need tools they can explain to banks, investors and board members. - eFinancialModels said transparent logic and visible assumptions make every number easier to defend in funding conversations. - eFinancialModels offers more than 3,100 financial model templates across sectors including real estate, SaaS, healthcare and manufacturing. - The templates are designed for immediate use with customizable assumptions, professional formatting and comprehensive documentation. - The company also points users to resources such as “7 Key Tips for Working With Financial Model Templates.” - The platform serves startup founders preparing pitch decks and investment bankers conducting valuations. - Template offerings range from free basic models to premium solutions. - Common use cases include discounted cash flow valuation, leveraged buyout analysis, startup financial planning and industry-specific operational modeling. - The source text includes links to eFinancialModels social accounts on LinkedIn, Bluesky, Instagram, Facebook, YouTube, TikTok and X.

Between the lines: - The article frames AI not as a replacement for model builders, but as a tool that still needs human verification and governance. - For regulated users, auditability appears to matter more than speed or novelty. - Downloadable templates are benefiting from that caution because they provide structure without obscuring the logic behind the numbers. - The competitive edge in financial modeling may be shifting from automation alone to explainability plus workflow fit.

What's next: - Industry observers expect the gap between AI capability and deployability to narrow as tools improve. - The Forvis Mazars report says modeling expertise still needs active maintenance, not a one-time purchase. - For practitioners, the immediate focus is tooling that supports current work while reinforcing long-term skills. - Professionals looking for template-based models can explore industry-specific options at eFinancialModels.com.