When Michael Tannenbaum, CFO of fintech company Brex, posted on X that the company’s AI agent now processes 99% of expense reports without any human involvement, the reaction from the tech and finance communities was swift and telling. The post, shared by Brex co-founder and CEO Pedro Franceschi’s colleague Michael Truell on X, highlighted a statistic that should give every back-office worker and enterprise software executive reason to pay close attention: artificial intelligence is no longer a theoretical threat to white-collar workflows — it is actively replacing them at scale.
Brex, the corporate card and spend management platform valued at over $12 billion, has been building AI capabilities into its products for the past two years. The company, which serves tens of thousands of businesses from startups to enterprises, has positioned itself at the intersection of financial services and artificial intelligence. The 99% automation figure represents a remarkable achievement in a domain — expense management — that has historically been defined by tedious manual review, policy enforcement headaches, and endless back-and-forth between employees and finance teams.
From Manual Review to Machine Autonomy: How Brex Got to 99%
The traditional expense report process is one that virtually every working professional has encountered and few have enjoyed. An employee submits a receipt, a manager reviews it, a finance team member checks it against company policy, and eventually — often after days or weeks — the expense is approved or flagged. The process is rife with friction, errors, and wasted time. According to a report from the Global Business Travel Association, the average expense report takes 20 minutes to complete and costs $58 to process. When errors are involved, that figure jumps to $52 per correction.
Brex’s approach to eliminating this friction has been to deploy AI agents — autonomous software systems that can interpret receipts, match them against corporate spending policies, detect anomalies, categorize expenses, and approve or flag reports without a human ever touching the process. The 99% figure suggests that only the most unusual or edge-case expenses now require human review, a level of automation that would have seemed implausible even three years ago.
The Broader AI Agent Movement in Enterprise Finance
Brex is not operating in isolation. The push toward AI agents in enterprise finance has accelerated dramatically in 2025. Companies like Ramp, which competes directly with Brex in the corporate card space, have also been investing heavily in AI-powered automation. Ramp has publicly discussed its own AI capabilities for automating receipt matching, policy enforcement, and accounting workflows. Meanwhile, established players like SAP Concur and Expensify have been racing to integrate large language model capabilities into their platforms to keep pace with the newer entrants.
What distinguishes the current wave of AI automation from earlier attempts at rules-based expense management is the flexibility and intelligence of the underlying systems. Previous generations of software could enforce simple rules — “no expenses over $500 without manager approval” — but struggled with the ambiguity and context-dependence that characterize real-world corporate spending. Modern AI agents, built on large language models and trained on vast datasets of financial transactions, can interpret context, understand policy nuance, and make judgment calls that previously required a human finance professional.
What 99% Automation Actually Means for Finance Teams
The implications of Brex’s achievement extend well beyond the company itself. If AI agents can handle 99% of expense reports autonomously, the question becomes: what happens to the people who used to do that work? The answer is nuanced. For many finance teams, expense report review has always been a low-value, high-volume task that consumed disproportionate amounts of time. Freeing those professionals from the drudgery of receipt checking could allow them to focus on higher-order work — financial planning, strategic analysis, and business partnering.
But the optimistic framing only goes so far. For organizations that employed dedicated accounts payable clerks or expense auditors, the math is stark. If a system can process 99% of volume without human involvement, the headcount required to manage the remaining 1% is a fraction of what was previously needed. This is the kind of concrete, measurable displacement that economists and labor researchers have been warning about — not in some distant future, but right now, in 2025.
The Trust Question: Can Companies Really Let AI Approve Spending?
One of the most significant barriers to AI adoption in financial workflows has always been trust. Money is involved, and errors — whether from fraud, misclassification, or policy violations — carry real financial and regulatory consequences. The fact that Brex has reached 99% automation suggests that the company has built sufficient confidence in its AI systems to let them operate with minimal human oversight.
This raises important questions about accountability and governance. When an AI agent approves an expense that turns out to be fraudulent, who bears responsibility? When a system misinterprets a policy and approves spending that violates compliance requirements, what recourse does the company have? These are not hypothetical concerns. As AI agents take on more autonomous decision-making authority in financial processes, the legal and regulatory frameworks governing their use will need to evolve in parallel. The SEC and other financial regulators have begun examining the use of AI in financial services, though much of the current focus has been on trading algorithms and advisory services rather than back-office automation.
Brex’s Strategic Position in the AI-Powered Finance Stack
For Brex, the 99% figure is as much a marketing statement as it is a technical achievement. The company has been locked in fierce competition with Ramp, Navan, and other corporate spend management platforms. In a market where the underlying financial products — corporate cards, expense management, bill pay — are increasingly commoditized, AI capabilities represent a meaningful differentiator. If Brex can credibly claim that its platform eliminates virtually all manual work from the expense process, that is a powerful selling point for CFOs and finance leaders evaluating vendors.
The company has also been expanding its AI capabilities beyond expense management. Brex has introduced AI-powered features for budget management, vendor payment processing, and financial reporting. The vision appears to be a fully autonomous finance back office, where AI agents handle the transactional and compliance-heavy work while human professionals focus on strategy and decision-making. It is a vision shared by many in the fintech industry, but Brex’s willingness to put a specific number on its automation rate — 99% — sets it apart from competitors who speak in vaguer terms about AI-assisted workflows.
The Workforce Implications Are Real and Immediate
The broader workforce implications of this kind of automation deserve serious examination. According to the Bureau of Labor Statistics, there are approximately 1.5 million bookkeeping, accounting, and auditing clerks in the United States. Many of these roles involve exactly the kind of transactional, rules-based work that AI agents are now capable of performing. While not all of these positions will be eliminated — many involve tasks beyond expense processing — the direction of travel is clear.
McKinsey & Company has estimated that generative AI could automate activities that currently absorb 60 to 70 percent of employees’ time in certain occupations. Finance and accounting roles are consistently identified as among the most susceptible to automation. The Brex example provides a concrete data point to support these projections. When a single company can automate 99% of a specific workflow, it is reasonable to expect that similar levels of automation will spread to adjacent processes and industries.
What Comes Next for AI in Corporate Finance
The trajectory from 99% automation of expense reports to broader autonomous financial operations seems inevitable. The technology is already being applied to accounts payable, accounts receivable, procurement, and financial close processes. Companies like Vic.ai, Stampli, and Tipalti are building AI-powered automation into various segments of the finance function, and the major enterprise resource planning vendors — SAP, Oracle, and Microsoft — are integrating AI capabilities into their platforms at an accelerating pace.
For finance professionals, the message from Brex’s announcement is both a warning and an opportunity. The roles that consist primarily of processing, reviewing, and approving transactions are being automated at a pace that few anticipated. The roles that involve judgment, strategy, relationship management, and creative problem-solving remain, for now, beyond the reach of AI agents. The professionals who recognize this shift and adapt accordingly will find themselves well-positioned. Those who do not may find that the 99% figure applies not just to expense reports, but to their job descriptions as well.
Brex’s achievement is a signpost, not an endpoint. The automation of corporate finance is accelerating, and the companies and professionals who understand what that means — and act on it — will define the next era of enterprise operations.
WebProNews is an iEntry Publication