This Week in Finance Automation

Week of February 22, 2026

Spend Matters published a three-part series this week asking a question most AP vendors don't want to hear: why does automation stop working when you try to scale it? The answer — only 32.6% of invoices are processed without human intervention — puts the entire industry's marketing claims in perspective. Meanwhile, Sage Intacct launched AI agents for mid-market finance teams, and a new IDC survey confirmed what we've been saying: accountants want AI to do the work, but they're not ready to let it make the decisions.

The Big Story

AP Automation's Scaling Problem: Why Early Wins Don't Compound

Spend Matters published a multi-part investigation this week examining a pattern most AP vendors won't talk about: automation that works beautifully for the first 100 invoices quietly breaks at 1,000.

The core finding is uncomfortable. Despite years of vendor promises, only 32.6% of invoices are processed without any human touch — what the industry calls "touchless" or "straight-through" processing. The rest hit exceptions, require manual review, or get stuck in approval limbo. The gap between the demo and reality grows wider as organizations add more suppliers, currencies, regions, and spend categories.

The series argues the problem isn't technology — it's architecture. Most AP automation tools are designed for happy-path processing. They handle standard invoices from known vendors with matching POs efficiently. But the exception rate climbs as complexity increases, and organizations that scaled successfully had to fundamentally redesign how they handle validation, ownership, and decision-making — not just bolt on more automation.

Source: Spend Matters

Our Take: This validates exactly why AI-native AP tools exist. Traditional rule-based automation hits a ceiling because rules can't adapt. AI-powered exception handling — where the system learns from each resolution — is the only architecture that improves with scale instead of degrading. If your AP tool gets worse as you grow, the tool is the problem.

Notable Developments

Sage Intacct Ships AI Agents for the Mid-Market

Sage announced its 2026 R1 release of Sage Intacct with several AI features that matter for AP teams. The headline is a Finance Intelligence Agent that lets controllers query financial data in plain English — think "show me all overdue invoices from Q4 vendors" instead of building custom reports. An AI-powered Import Agent simplifies data migration with natural language transformations and flexible mapping. New Customer Payment Services (powered by Fortis) add instant settlement with automated reconciliation.

When the dominant mid-market ERP ships native AI agents, it signals that agentic AI in finance is moving from "innovative" to "expected." AP tools that don't offer AI-powered querying and automation will feel outdated by year-end.

Source: Sage

Inscope Raises $14.5M as AI-Native Finance Tools Keep Winning

Inscope, an AI-powered financial reporting platform, closed a $14.5M Series A led by Norwest. The numbers are striking: customer base grew 5x and annual recurring revenue increased 30x over the past twelve months. Founded by two former Flexport finance leaders, the platform automates financial statement preparation, review, and delivery.

The 30x ARR growth signals genuine demand for AI tools that replace manual finance workflows, not just augment them. Investors are betting that the same wave automating AP and bookkeeping will reshape financial reporting.

Source: TechCrunch

Two-Thirds of Accountants Use AI — But 88% Insist on Human Oversight

A new IDC study surveying 1,005 accounting professionals across six countries found that 66% already have AI embedded in their firm strategy or active pilots underway. But the trust gap remains wide: 88% say AI carries risk of undermining professional judgment, and 64% believe validation of AI outputs should always be required.

The top barriers? Cost of implementation (34%) and lack of technical talent (30%). Despite the caution, 76% believe AI will fundamentally transform accounting within a decade.

Source: GlobeNewswire

Quick Hits

Alphabet: CFO Anat Ashkenazi confirmed Google's finance team uses agentic AI for invoice payment and reconciliation. About half of Google's code is now AI-generated.
UiPath + WorkFusion: UiPath completed its acquisition of WorkFusion, adding pre-built AI agents for AML, KYC, and sanctions screening to its automation platform.
AI Native Accounting Foundation: A new nonprofit launched to help accounting professionals separate AI hype from practical implementation. Awards program open until April 15.
Forrester: Identified 41 AP invoice automation vendors and four 2026 trends: e-invoicing compliance, data quality challenges, agentic AI as differentiator, and intelligent invoice-lifecycle networks.
NY FAIR Act: New York's first consumer protection law update in 45 years took effect Feb 17, expanding "deceptive" to include "unfair" and "abusive" business practices.

Numbers of the Week

Metric	Value	Context
Touchless invoice rate	32.6%	Share of invoices processed without human intervention — the real benchmark for AP automation (Spend Matters)
AI adoption in accounting	66%	Accounting professionals with AI embedded in strategy or active pilots (IDC/Caseware)
Inscope ARR growth	30x	Year-over-year recurring revenue growth for the AI reporting platform (TechCrunch)
AI oversight demand	88%	Accountants who say AI risks undermining professional judgment (IDC/Caseware)

What We're Watching

The Spend Matters series on AP scaling failures is the most important conversation happening in AP automation right now. If the industry's best-case touchless rate is 32.6%, there's a massive opportunity for AI-native tools that handle the other 67.4% — the exceptions, the edge cases, the invoices that don't match a PO, the vendors with inconsistent formats.

Sage Intacct's AI agent launch is also worth tracking. When the mid-market's dominant ERP adds native AI querying, it raises the bar for every tool in the stack. AP automation solutions that can't answer plain-English questions about payment status, vendor spend, or approval bottlenecks will feel increasingly behind.

And the IDC survey's 88% "insist on human oversight" number should be tattooed on every AI vendor's product roadmap. The winners won't be the tools that promise full autonomy — they'll be the tools that make humans faster, more accurate, and more confident in their decisions.

The Bottom Line

The gap between AP automation's promise and its reality is finally getting honest scrutiny. Only a third of invoices are truly touchless. Two-thirds of accountants use AI but nearly nine in ten demand human oversight. The industry is moving from "AI can do everything" to "AI should do the right things." That's progress. Tools that embrace this nuance — automating the tedious while preserving human judgment where it matters — will define the next generation of finance automation.

Subscribe to get This Week in Finance Automation delivered every week.

This Week in Finance Automation: AP Hits a Scaling Wall | Week of Feb 22, 2026