AI in ERP: SAP, Dynamics 365, and Workday for Finance

Before finance teams go looking for external AI tools, there is a question worth asking: how much AI does your ERP already contain? The answer, for organisations running SAP, Dynamics 365, or Workday, is: probably more than you think. All three major ERP platforms have been embedding AI capabilities into their finance modules for several years, and the pace of investment has accelerated significantly since 2023.

The problem is that most organisations are not using these native AI features. Either the finance team does not know they exist, IT has not activated them, or the organisation is running an older version that does not include them. This guide cuts through the vendor marketing to give you a clear picture of what each ERP actually offers for finance, what it genuinely does well, and how to get started activating capabilities you may already be paying for. For broader context, see our guide to AI use cases in finance.

Your ERP Already Has AI

SAP, Microsoft, and Workday have each invested billions in embedding AI into their platforms. This investment shows up in very different ways across the three ERPs - different terminology, different feature architecture, and different strengths - but the common thread is that meaningful AI capabilities are available within the ERP environment that most large finance teams already operate.

The case for prioritising native ERP AI over external tools is strong: the data is already in the system, so there is no extract-transform-load overhead; the security boundary does not extend to external services; the AI operates in the context of your actual ERP data rather than copies or samples; and the integration with ERP workflows is seamless rather than requiring custom connectors. For most finance tasks, native ERP AI is simpler, more secure, and more deeply integrated than any external tool.

The case for external tools alongside native ERP AI is also real: external tools often have more powerful general capabilities, particularly for text analysis, commentary drafting, and creative financial thinking; they update faster than ERP software release cycles; and they are accessible to finance team members for tasks that do not involve ERP data. The best approach for most organisations is native ERP AI for in-system tasks and external tools like ChatGPT, Copilot, or Claude for analysis and communication tasks outside the ERP.

SAP AI for Finance

SAP's AI strategy in finance is built around three pillars: Joule (the conversational AI assistant), Business AI embedded capabilities, and the broader SAP Business Technology Platform (BTP) which provides the infrastructure for custom AI applications.

SAP Joule

Joule is SAP's generative AI assistant, embedded across S/4HANA Cloud and other SAP applications. For finance teams, Joule provides natural language access to ERP data - you can ask questions like “show me the top 10 overdue accounts by value” or “what is the variance between actual and planned OPEX for Q3?” and Joule retrieves the relevant data and presents it in a structured format.

Joule also assists with process guidance - walking users through SAP transactions in natural language, reducing the training burden for less experienced users and improving efficiency for experienced users who do not need to navigate menu structures for routine queries. For finance teams with SAP S/4HANA Cloud, Joule is included and can be activated without additional licensing in most configurations.

Predictive Accounting

SAP's predictive accounting capability uses machine learning to forecast the financial impact of operational events before they are formally recorded. When a sales order is created, for example, predictive accounting can estimate the expected revenue recognition timing, the likely payment date based on customer behaviour, and the impact on cash flow - giving finance teams a real-time forward view rather than a backward-looking record of completed transactions.

This capability is particularly valuable for organisations with complex revenue recognition rules (IFRS 15 or ASC 606 compliance) and for treasury teams that need accurate short-term cash flow forecasting. Predictive accounting learns from historical data and improves accuracy over time as it processes more transactions from your specific customer base.

Automated Cash Application

SAP's intelligent cash application uses AI to automatically match incoming payments to open receivables, even when payment references are incomplete, partial, or missing. Traditional cash application is a significant source of manual effort in high-volume AR environments. SAP's AI model learns from your historical matching patterns and improves its matching accuracy over time, reducing manual matching rates significantly in most implementations.

Dynamics 365 Copilot for Finance

Microsoft has integrated Copilot more deeply and visibly into Dynamics 365 Finance than any other ERP vendor has integrated AI into their platform. This reflects Microsoft's overall AI strategy - embedding Copilot across every Microsoft product - and gives D365 Finance users access to generative AI capabilities that are natively connected to their ERP data, their Microsoft 365 ecosystem, and the broader Microsoft Azure AI infrastructure.

Copilot in D365 Finance

Copilot in Dynamics 365 Finance provides natural language querying of financial data, automated report generation, and AI-assisted process completion. Finance users can ask Copilot to pull specific reports, explain variances, summarise account activity, or guide them through complex transactions - and Copilot accesses live ERP data to provide accurate, contextually relevant responses.

The Copilot integration also extends into the broader Microsoft 365 ecosystem. Finance data from D365 can be surfaced in Excel, Word, and Teams through Copilot, allowing finance professionals to query ERP data from within the Office applications where they do their analysis and reporting work. This cross-application integration is a significant productivity multiplier for teams that move between D365 and Office regularly.

Customer Payment Predictions

D365 Finance includes an AI-powered customer payment prediction feature that uses machine learning to forecast when individual customers are likely to pay their outstanding invoices. The model learns from your historical payment data and considers factors including customer payment history, invoice amount, invoice age, and broader payment behaviour patterns.

This capability has direct financial impact. When you know which invoices are likely to pay late, your collections team can prioritise their outreach more effectively. When treasury knows which customer payments are likely to come in on time versus late, cash flow forecasting accuracy improves significantly. D365 Finance presents these predictions at the customer and invoice level within the standard accounts receivable workspaces, making them accessible without any special training.

Process Automation

D365 Finance includes native AI capabilities for process automation across the procure-to-pay and order-to-cash cycles. Intelligent invoice capture reads and extracts data from invoices automatically, matching to purchase orders and flagging exceptions. Vendor invoice matching uses AI to handle three-way matching at scale, reducing the manual effort involved in AP processing for high-volume environments.

Workday AI for Finance

Workday's AI strategy is built into its Workday AI platform, which powers ML-driven capabilities across the Human Capital Management (HCM) and Financial Management applications. For Workday Financial Management users, the most impactful AI capabilities are in planning, anomaly detection, and automation.

Workday Adaptive Planning with ML

Workday Adaptive Planning is Workday's dedicated financial planning and analysis application, and it includes sophisticated ML forecasting capabilities that go significantly beyond traditional spreadsheet-based or deterministic planning models. The ML engine analyses historical patterns, identifies seasonality and trend factors, and generates statistical forecasts with confidence intervals - giving FP&A teams a more rigorous starting point for their planning cycles.

The ML forecasting in Adaptive Planning is particularly strong for revenue forecasting in organisations with large volumes of historical transaction data - SaaS businesses, financial services, retail, and subscription-based models. The models learn from your specific data patterns rather than generic industry benchmarks, which means accuracy improves over time as the models accumulate more historical data to learn from.

Anomaly Detection

Workday Financial Management includes AI-powered anomaly detection that continuously monitors financial transactions and flags unusual patterns - potential duplicate payments, transactions outside policy, unusual vendor activity, or statistical outliers in spending patterns. These alerts surface in the Workday interface and can be routed to the appropriate reviewer automatically based on the nature and value of the anomaly.

Unlike traditional rule-based controls that can only catch known patterns, ML anomaly detection learns what “normal” looks like for your specific organisation and flags deviations from that baseline. This means it can surface novel fraud patterns or control failures that rule-based systems would miss entirely.

Skills Cloud for Finance

Workday's Skills Cloud is primarily an HCM capability, but it has direct finance team implications. It uses ML to build a comprehensive skills taxonomy across your workforce, allowing finance leaders to understand current AI and digital skills across their finance team, identify skills gaps for AI adoption initiatives, and plan training interventions accordingly. For CFOs planning AI adoption programmes, Skills Cloud provides data on where their team currently stands and what skills are needed.

Third-Party AI Add-Ons

Beyond native ERP AI, a specialist ecosystem of third-party AI applications integrates with SAP, D365, and Workday to provide deeper capabilities in specific finance domains. Three categories are particularly established in 2026.

HighRadius provides AI-powered order-to-cash automation - covering credit risk assessment, collections management, cash application, and deduction management - that integrates with all three major ERP platforms. Its AI models are trained on large datasets of financial transactions, giving it pattern recognition capabilities that individual ERP implementations cannot match. For organisations with high-volume AR operations, HighRadius typically delivers faster DSO reduction than native ERP AR features alone.

BlackLine specialises in financial close automation, with AI capabilities covering reconciliation matching, journal entry validation, and inter-company accounting. It integrates natively with SAP, D365, and Workday, sitting between the ERP and the finance team to automate the close-related tasks that are most labour-intensive. BlackLine's AI learns from historical reconciliation patterns to improve matching accuracy over time.

Trintech offers similar financial close capabilities with strong SAP integration. Its AI-powered reconciliation and close management tools are particularly well-established in financial services and insurance, where the volume and complexity of period-end reconciliations is particularly high.

How to Activate ERP AI

Getting value from ERP AI requires more than simply knowing the features exist. Many organisations have AI capabilities sitting unused in their ERP because activation requires deliberate configuration, data quality work, and change management. Here is how to approach activation systematically.

Step 1: Audit what you have. Request a capability review from your ERP vendor or implementation partner. List every AI feature available in your current version and licence tier, and identify which are activated versus available but inactive. This audit often reveals significant untapped capability.

Step 2: Assess data quality. AI capabilities in ERP systems are only as good as the underlying data. Customer master data, vendor data, and transaction history must be sufficiently clean and complete for AI models to learn effectively. Prioritise data quality remediation for the AI features you intend to activate first.

Step 3: Prioritise by impact. Use the FAIR framework to evaluate which native AI features to activate first. Start with features that score highly on Fit (they address a current pain point) and where Accuracy is verifiable (you can check outputs against known correct answers).

Step 4: Train your finance team. Native ERP AI features are underutilised partly because finance teams do not know they exist, and partly because they do not know how to use them effectively. Training on ERP-native AI features - including Joule for SAP, Copilot for D365, and Workday's AI capabilities - is now part of a comprehensive AI skills development programme for finance teams.

Step 5: Govern and monitor. As with all AI deployments in finance, ERP AI features require governance - particularly around who can initiate automated actions, what the audit trail requirements are, and how AI-generated outputs are validated before use in financial reports. Our AI for Finance Leaders course includes a dedicated module on AI within ERP systems, covering SAP, D365, and Workday. Our D365 consulting team and Workday consulting team can support AI feature activation and configuration as part of broader ERP optimisation engagements.

Frequently Asked Questions