Agentic Finance AI Gains Traction Amid Governance Gaps and $23M Investment Surge

Despite soaring investment and technological innovation, the deployment of agentic finance AI remains largely experimental across global enterprises, raising urgent questions about governance, scalability, and measurable return on investment. According to a recent FT Longitude survey of 200 senior finance leaders in the US, UK, France, and Germany, 61% have deployed AI agents only as pilot projects, while one in four admit they lack clear metrics to evaluate success. This disconnect between ambition and execution underscores a critical gap in enterprise AI strategy—even as venture capital pours into startups promising to bridge it.

One of the most compelling signals of market confidence comes from Stacks, a London-based agentic AI platform for enterprise finance, which announced a $23 million Series A round on February 19, 2026. Led by Lightspeed, with participation from EQT Ventures, General Catalyst, and S16VC, the funding round more than doubles the company’s total capital since its $12 million seed round just 11 months prior. Stacks’ platform, which orchestrates autonomous AI agents to automate complex finance workflows—from invoice reconciliation to cash flow forecasting—positions itself not as a tool, but as an operational layer that replaces manual decision-making with dynamic, goal-driven AI agents. "We’re not building chatbots or predictive models," said Albert Malikov, CEO and co-founder of Stacks. "We’re building digital finance employees that own outcomes, not just tasks."

Investors are betting that agentic AI—systems capable of autonomous planning, reasoning, and execution—will finally deliver on the long-promised efficiency gains in finance operations. Unlike traditional AI that requires human prompts, agentic systems can initiate actions, adapt to changing conditions, and collaborate with other agents. For example, a Stacks agent might detect an anomaly in vendor payments, investigate historical patterns, negotiate a discount with the supplier via integrated communication channels, and update the company’s cash flow model—all without human intervention. This level of autonomy, if properly governed, could reduce month-end close cycles by up to 70% and cut fraud-related losses by 40%, according to internal case studies cited by the company.

Yet the FT Longitude findings reveal a sobering reality: most organizations lack the governance frameworks to safely scale such autonomy. Without clear accountability protocols, audit trails, and ethical guardrails, agentic AI risks generating erroneous decisions, compliance violations, or even financial losses. "We’ve seen CFOs deploy AI agents to optimize working capital, only to find them liquidating inventory too aggressively during seasonal dips," said a former finance director at a Fortune 500 retailer who requested anonymity. "The AI was optimized for speed, not sustainability."

Stacks attempts to mitigate these risks through its "Trust Layer," a built-in governance architecture that logs every agent decision, enforces role-based permissions, and integrates with existing ERP systems like SAP and Oracle. The company also partners with enterprise risk consultancies to embed compliance checkpoints aligned with SOX, GDPR, and local financial regulations. "You can’t have agentic AI without agentic accountability," Malikov emphasized.

The convergence of venture capital momentum and enterprise caution suggests a pivotal moment for agentic finance AI. While startups like Stacks race to build scalable platforms, corporate finance departments must catch up with policy, training, and oversight. Without structured governance, the $23 million investment in Stacks may be just one drop in a sea of unfulfilled potential. As global CFOs face mounting pressure to deliver immediate ROI, the next 18 months will determine whether agentic AI becomes a transformative force—or another expensive experiment.