Skip to main content

Interviews conducted by Forrester suggest that artificial intelligence (AI) has yet to help transform finance, but is advancing rapidly in some areas.

Procure-to-pay (P2P), for example, uses natural language processing (NLP) and machine learning (ML) and has shown immediate returns, while analysis of cash ordering and audit shows the short-term benefits of AI. Additionally, predictive analytics can augment basic Business Intelligence (BI) reporting for financial planning.

Four Ways AI Empowers Finance and Accounting

Audit analytics, procure-to-pay, order-to-pay, and financial planning are four finance and accounting (F&A) processes where the AI ​​technology needed to elevate the process already exists. There is also an active community of technology vendors, and customer references indicate solid progress. Forrester gives these four use cases strong adoption scores, such as a manageable skills gap, stable data, and clear business results.

Analytical audit

Starting with audit analysis, auditors tend to spend too much time buried in compliance checklists and creating reports that few people read, with little time to look for anomalies in every transaction. Rather than manually sampling data points, Forrester says machine learning is used for transaction risk assessment.

The industry member association American Institute of Certified Public Accountants (AICPA) develops guidelines for CB in the audit function. Mature audit support providers like Thomson Reuters and Wolters Kluwer, as well as emerging companies like Caseworks Cloud and MindBridge, are integrating AI into their audit platforms.

Technology readiness is high, with mature ML, while NLP mining involves unstructured content such as emails. The adoption profile is also strong, with few governance issues, high business value, and high potential for disruption. However, the training of auditors in ML aspects reveals a current skills gap.

Obtain to pay

Examining the Procure-to-Pay (P2P) process, Forrester found that P2P can leverage ML to standardize and analyze spend, contract, market, and vendor data. Augmented BI can isolate payments that once incurred late fees and surface billing exceptions, categorize expenses for tracking, onboard new vendors faster, and automatically detect fraud.

One important area is invoice processing, where Level 1 capture, optical character recognition (OCR), and workflow automation models have been applied for decades. Early solutions were template-based, where the extraction rules aligned with a specific invoice or purchase order template. New approaches use NLP to provide model-free and area-free mining. To guarantee the quality of the digitization, each extraction can be stamped with a level of certainty.

ML can handle complex document structures more easily, without pre-configuration. Both NLP and traditional ML are mature, offering a strong technology readiness score. The adoption profile is also strong due to high potential for disruption and a mix of stable data from semi-structured forms and data.

Order to cash

Order to cash is another untapped candidate for AI-based automation. Cash is the lifeblood of most businesses, yet it remains underserved by the latest automation practices, especially in relation to P2P.

In most cases, Accounts Receivable (AR) invoice automation software generates the customer invoice in formats such as CXML (Commerce XML), ebXML (E-Commerce XML), and Edifact, and tracks the status, while F&A manages the money. Modern order-to-cash solutions elevate the role of the AR professional as many tasks are shifted to AI-powered bots that can take over email communications or create a collection letter based on the automatic classification, the basic system data and the stage of the dispute.

Analytics will control cash applications. AI will drive automated payment lifecycles, credit management and predictive forecasting of remittances. The adoption profile is strong due to a clear business outcome such as improved cash performance. Rules-based workflows and decisions are starting to give way to AI-based ones, but the business value is now moderate. The technology is ready today, with ML, robotic process automation (RPA), and text analytics ready to help.

Financial planning

A fourth use case, financial planning and analysis, is starting to go beyond Excel. Financial analytics has great potential for AI support, but most financial services depend on Excel or basic reporting from specialist vendor platforms.

However, future planning and budget forecasting will use simulation, optimization, and ML-based statistical modeling that links business strategy to execution. An example is Vena Solutions, which offers a Microsoft-oriented M&A product with built-in Power BI to provide an easy path to predictive analytics and machine learning (PAML).

Four areas where AI in finance and accounting still needs to be developed

Contract analysis

Contract analytics has broad potential in a number of use cases. Contract analysis is not a core finance and accounting function, but it is increasingly of interest to chief financial officers (CFOs) and their staff.

The primary use of AI is to automate the import and markup of legacy and third-party contract metadata. Platforms like ContractPodAI and Icertis and specialist AI vendors like Corticol.io are integrating AI functions into contract lifecycle management (CLM).

ML can help assess risks and anomalies across the contract portfolio, find contracts with wording related to a new issue or topic such as Brexit or new tax laws, and populate CLM or d other workflow automation platforms to support service level agreements (SLAs) and other terms and conditions. Forrester sees early progress in providing chatbots to help put together a contract project.

The main building block of the AI ​​is text analytics, which provides a moderate to high technology readiness score. Forrester’s Adoption Profile score shows high business value but unclear results, and variety of document formats makes data less stable.

Reconciliation of accounts

AI can also be used in account reconciliation to resolve data issues. Many tasks in finance and accounting require two or even three sets of documents to agree, especially when money leaves a bank account. Prepaid expenses, bad debts, fixed assets, cash accounts, and general ledger and subledger ledger tasks are typical targets to reconcile. Missing or lost transactions, unreconciled accounts, or improper use of postings are typical.

ML can handle a wide variety of structured data sources in many formats (CSV, XML, SQL or NoSQL) where it can “learn” data sources and patterns, with data control rules in a central location . Most reconciliations only deal with two sources of data, but AI can extend this to multiple sources. RPA bots help extract data, provide data entry support, and run an approval process.

In the account reconciliation, Forrester notes that the technology maturity is high, but the adoption profile is medium due to low data stability and moderate business and disruption value. Reconciliation is often a necessary secondary function of the closing process.

Automated closing

The automation of monthly and quarterly closes is based on a basic workflow. Automating the closing process is a priority for many M&A departments. A well-run fence is a sign of a well-run business. Transparency, timeliness, accuracy and timeliness of reporting are key concerns. Nearby automation should integrate with business applications, spreadsheets, and various accounting systems to document relevant data and identify inconsistencies.

AI has the potential to collect data from different sources, bring it together and merge it, speed up the monthly process and be more accurate. Automation focuses on checklists and individual tasks, tracking the closing process, deadlines, and approvals.

Forrester rates closing the books with an average technology readiness score but a high adoption profile, driven by high business value and disruptive potential – although core system acquisitions and conversions make the data less stable.

Expense management

Another area where AI and automation can be used is expense management. Forrester hasn’t seen a big push to use advanced forms of AI in this area. RPA-centric intelligent automation was the main improvement. For example, a federal agency is using RPA bots to verify line item details, which humans used to do. When travel resumes, this back-office function will once again become an essential part of any well-run business. Major market players like SAP Concur haven’t pushed AI, nor have CFOs looking for efficiency or anomaly detection.

Forrester rates technology readiness for using AI in expense management as high due to reliance on RPA bots and traditional ML. However, despite clear results, the adoption profile is lower due to low perceived business value, minimal potential for disruption, and unstable data.

In summary, Forrester notes that many finance and accounting processes have unnecessary variation. The AI ​​works best against standard and repeatable actions. But extra process steps, offline behaviors, malicious spreadsheets, and personal shortcuts are common. This lack of standardization of tasks across companies prevents software vendors from creating focused, easy-to-implement AI for financial and accounting processes.


This article is based on an excerpt from the Forrester report “AI in finance and accounting”. Craig Le Clair is vice president and principal analyst at Forrester.