Oracle and NetSuite are moving AI toward business execution, but the harder test is whether customers can adopt it, govern its decisions, and improve operations long after go-live.
Oracle AI World and Oracle NetSuite SuiteWorld 2026 showed how far AI has advanced and how much adoption work remains. Oracle emphasized connected data, infrastructure, and applications, while NetSuite focused on agents, daily workflows, and implementation. The question is whether customers can put those capabilities to work and keep improving after go-live.
In my Forbes article, ERP Innovation Is Moving Faster Than Customer Adoption, I examined why ERP modernization continues after go-live. Data quality, process ownership, and employee readiness all affect what customers get from ERP. Connecting implementation knowledge with support and customer success can help close that gap. Someone needs to own progress after implementation, and vendors need to understand what customers face in production.
AI value does not arrive alongside ERP. It arrives through it. The question is how much easier Oracle and NetSuite make it for customers to use what they already have.
Oracle’s AI Strategy Depends on Trusted Context and Decision Authority
Oracle’s expanded OpenAI collaboration brings AI models closer to Oracle applications and Oracle Cloud Infrastructure (OCI). Oracle AI Workspace is designed to connect business questions with approved work across Fusion Applications, NetSuite, and other systems. That could save employees from moving between applications to get work done. But being able to see a record is different from having permission to act on it.
Consider a finance manager investigating a delayed payment. Receivables, customer commitments, and credit policies may sit in different systems. An agent could assemble the facts and recommend a credit hold. Whether the agent should place the hold depends on the policy, the account relationship, and the person authorized to approve the decision.
Oracle’s AI Data Platform is part of the answer. Its proposed ontologies and semantic reasoning could help agents interpret business relationships and policies rather than treat data access as sufficient context. Trusted context is the currency for enterprise AI. Someone needs to keep that context current as business rules, processes, and systems change. Even accurate data can lead to the wrong action if decision ownership is unclear or an agent cannot determine which policies, approvals, or exceptions take precedence across systems.
Oracle’s plans for Google tensor processing units (TPUs) on OCI starting in 2027 and Snowflake as a native OCI service offer more options. Customers already have enough complexity. More technology choices help only when they work together. Oracle should show how it resolves conflicting data, preserves decision-making authority, and addresses data and process gaps before agents take on more responsibility.
NetSuite’s Autonomous Vision Meets the Reality of ERP Operations
At SuiteWorld, Evan Goldberg, founder and executive vice president of Oracle NetSuite, outlined an autonomous suite where NetSuite Next, Ask Oracle, SuiteSkills, AI Workspaces, and SuiteAgents can support decisions and coordinate approved actions. This builds on the direction I explored in my NetSuite Next research at SuiteWorld 2025. I see this as ERP’s third era, moving beyond transactions and visibility to business context, decisions, and governed action.
Goldberg described this shift as “the end of ERP as we know it,” meaning employees should be able to get answers and complete approved work without navigating complex screens. The way people use ERP may change, but the need for accurate records, controls, and human accountability does not.
NetSuite’s proposed Autonomous Close workspace could help finance teams address accounting exceptions throughout the month instead of waiting until close. A faster close matters only if the numbers are right, approvals can be traced, and finance remains accountable. NetSuite’s more than 45 SuiteSkills, available for use with Ask Oracle and connected AI services under existing permissions, could help employees use ERP capabilities they already have.
Trust will grow as these capabilities handle disputed reconciliations, exceptions that need a controller’s approval, and decisions that may need to be reviewed or reversed. Handoffs, thorough records, and reliable reversals will matter even more than another agent demonstration.
QuickStart and SuiteExpress Extend the ERP Journey Beyond Go-Live
QuickStart and SuiteExpress serve different stages of the customer journey. In my initial assessment, I liked that QuickStart did not ask startups to take on a full ERP system from day one. Its free-to-start offering for eligible U.S. technology startups covers purchasing, basic inventory, subscription billing, banking, and reporting within published usage limits.
Customers should be able to check an AI answer against the transactions behind it, understand when an approval is needed, and see usage limits before running into them. They also need clarity about what carries forward and what costs more when moving to full NetSuite. QuickBooks and Xero already offer familiar experiences; NetSuite’s opportunity is a clear path into ERP without importing full-suite complexity on day one.
SuiteExpress applies AI to requirements, migration, configuration, testing, and guided learning. NetSuite says suitable deployments could take weeks or even days, with availability for software-industry customers in the United States and Canada. Going live sooner helps. What matters afterward is whether employees and support teams can still use the requirements, configuration decisions, and process knowledge gathered during implementation.
Test results and repeat issues should also help vendors improve support and the product itself. I also analyzed Infor’s forward-deployed engineering approach, noting that lessons from deployments should reduce recurring problems and improve what subsequent customers can adopt.
Support teams see recurring issues and workarounds, while partners understand customer configurations and integration challenges. Customer success teams should bring that knowledge together with business owners to identify where training, data improvements, process changes, or additional capabilities are needed.
Oracle and NetSuite can use what they learn from implementations and support cases to improve partner guidance, product decisions, and customer adoption. A closed support ticket is not evidence that an adoption problem has been solved.
Microsoft Raises the Stakes for Enterprise AI Execution
Microsoft is pursuing a related path through Microsoft Fabric, OneLake, Azure, Copilot Studio, and Dynamics 365. In Microsoft’s Data Strategy Drives Autonomous ERP and my InfoWorld analysis of connected enterprise data, I examined how data platforms and applications can support decisions and execution rather than stopping at reporting.
Microsoft has an advantage because employees already use many of its tools throughout the day. Oracle brings database and application context, while NetSuite connects financial and operational processes for growing businesses. Customers will use a mix of tools. The system that owns a transaction still needs to enforce its rules.
As employees move from ChatGPT or Microsoft Copilot into enterprise applications, business context, permissions, and accountability must follow.
Owning the interface where work begins is not the same as having authority to complete a transaction.
Customer Outcomes Reveal Where AI Is Making a Difference
Construction is a good example of why business context matters. A payment application depends on verified job progress, approved changes, and contract terms, not simply faster document generation. This aligns with my research on industrial AI and event-driven agentic ERP, where operational context must shape what an agent recommends and can do. At SuiteWorld, NetSuite announced the rollout of a new construction accounting solution designed to address this challenge. While no customers are currently live on the solution, NetSuite expects customers to begin using the solution and seeing results in the coming months.
Other SuiteWorld customers offered more measurable evidence. Every Man Jack reported approximately $350,000 in annual cost savings, an estimated 4,500 hours saved annually, and a 50% reduction in financial close time. JonnyPops reported nearly halving its monthly close, while Terlato reported automating more than 100,000 order-status updates annually. These customer-reported results reflect connected ERP processes, automation and AI together; the newly announced agents have the potential to build on these results further.
Saving time is a start. The question is whether it helps customers reduce billing delays, improve cash flow, or make better inventory decisions. Support and customer success teams can use those results to guide future deployments and product improvements. Finance and operations leaders can also measure how long it takes to resolve an exception correctly.
ERP’s Next Test Begins After Go-Live
For Oracle, the test is whether business context and decision authority hold up across systems, especially when something goes wrong. For NetSuite, it is whether QuickStart, SuiteExpress, and agents help customers learn and use ERP, not just get it running faster. Both should distinguish what they can deliver today from what they plan to offer and show whether those improvements last over time.
These priorities build on the decision architecture I outlined in my Forbes analysis of Salesforce Headless 360, connecting systems of record, trusted context, decisions, governed actions, and measurable outcomes. Exceptions, employee decisions, and business results need to feed back into support, processes, and product development. That is how modernization continues after a project ends.
Over the next 12 to 18 months, I’ll watch whether Oracle and NetSuite close the gap between innovation and adoption as Microsoft pursues the same enterprise workflows. Neither an infrastructure portfolio nor a collection of agents will settle the competition. The vendors that stand out will be those helping customers solve operating problems, keep control of decisions, and build on their results after go-live.
Technology enables transformation. People determine the outcome.
