AI automation is already knocking on the door of ERP systems, and this is no longer a thing of the future. It is here to stay and is changing the way companies handle everyday tasks.
But along with this wave of innovation come very legitimate questions that companies need to answer before pressing any buttons:
- How is AI usage monitored within the system?
- What happens with sensitive company data during this process?
- How do you keep everything up to date while AI providers evolve at breakneck speed?
These are real concerns, and it makes total sense to raise them before any adoption. Nobody wants to start plugging AI into their processes without being sure everything is under control, right?
That is exactly the thinking behind Acumatica launching its AI Automation framework in version 2026 R1.
The idea is not just to automate for the sake of automating. The goal is to bring visibility, control, and governance into the ERP environment, allowing teams to adopt automation with confidence — without sacrificing consistency or data security.
In this article, you will learn how the feature works in practice, what is needed to activate it, and how organizations can manage it efficiently. 🚀
How AI automation works inside Acumatica
AI Automation allows Acumatica to connect to language model providers (the well-known LLMs) to help with tasks like generating responses, auto-filling fields, or summarizing information. And the most interesting part: all of this happens within controlled parameters based on well-defined rules.
When the feature is enabled and configured, AI automation adds a new command to the More menu. This command respects all existing access controls in the system, user permissions, and the audit log. In other words, no free-for-all: the AI operates within the same rules that already govern the rest of the ERP.
In practice, it works like this: when a user clicks that command to generate a prompt, the language model takes care of the rest. Simple to use, but with a robust structure behind the scenes ensuring everything is secure and traceable. This combination of simplicity on the front end and complexity behind the curtain is exactly what makes the tool so powerful for daily team workflows.
What you need before getting started
Before jumping into AI Automation, some fundamental requirements need to be in place. You really cannot skip these steps.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
- The AI Automation feature needs to be enabled through the Acumatica feature activation/deactivation screen, and it is subject to licensing.
- Your organization needs an account with a compatible LLM provider and an already deployed model.
- Internal data handling policies need to allow information to be sent to the provider for processing.
All this care around prerequisites exists for a good reason: it helps companies adopt AI intentionally, not reactively. Instead of chasing after technology just because it is trendy, the organization plans each step thoughtfully. That makes all the difference in the final outcome. 👌
Governance through well-defined roles
One of the biggest differentiators in Acumatica‘s approach to AI automation is the focus on role separation. Instead of giving everyone free rein to create prompts without oversight, responsibilities are clearly divided. This promotes consistency and supervision throughout the entire process.
It is worth highlighting an important point here: organizations should establish review processes appropriate to their business use case to validate AI-generated results before acting on them. AI is a great assistant, but the human eye remains essential to make sure everything is on point.
To keep all of this organized, two new out-of-the-box roles have been added to Acumatica. Let us break down what each one does. 👇
Prompt Engineer
This professional is responsible for:
- Creating prompt definitions
- Adding system instructions
- Testing prompts using real business data
Security Expert
This professional handles:
- Creating and maintaining system instructions
- Reviewing and approving prompts
- Ensuring alignment with security and compliance standards
This separation helps protect both data and outcomes. With this division of responsibilities, AI responses become more reliable, auditable, and suitable for enterprise use. It is the kind of structure that prevents headaches down the road. 🔒
Monitoring usage and token consumption
A really cool addition is the AI Automation History screen. It gives AI administrators a single place to review all artificial intelligence usage across their systems. Want to know when an action was executed? Need to check token consumption? Or trying to troubleshoot an error? The History screen has it all in one spot.
This screen offers detailed tracking of:
- Token consumption (both input and output)
- Errors and execution results
- The type of AI action that was performed
Each record indicates whether the activity was:
- A connection test
- A prompt test
- An AI-generated action on a form
This visibility gives administrators exactly the kind of information they need to fine-tune their strategy over time. For example, if output tokens are running too high, teams can choose a different LLM or refine prompts to make them shorter and more cost-effective. And if errors pop up, configuration or instruction updates can be made before problems snowball. That is real control at your fingertips. 📊
Staying current without waiting for platform updates
AI providers evolve at an impressive pace, adding or changing parameters all the time. Waiting for a full ERP platform update just to keep up with that pace is simply not practical. Nobody has that kind of time.
With that in mind, the LLM Connections feature in Acumatica helps administrators:
- Modify or create custom connection parameters
- Adjust provider settings directly
- Respond quickly to changes on the provider side
This flexibility is crucial in a market where AI providers shift gears constantly. It helps organizations stay up to date with AI innovations while remaining within a controlled and secure ERP environment. The best of both worlds, you know? 🌍
Building consistent and reliable AI prompts
The LLM Prompts screen is where consistency really matters. Instead of relying on improvised, on-the-fly instructions, prompts are defined with clear structure and purpose.
A well-designed prompt includes:
- Clear instructions
- Defined input data, sourced only from the active form
- Explicit output field definitions
Prompts also follow a standard instruction layout, which helps a lot with organization:
- Context instructions (tone, role, business context)
- Instructions with input data
- Output field definitions with their constraints
This structure promotes more consistent responses, faster prompt creation, and better alignment with business expectations. Worth mentioning: the Prompt Definition and the LLM command are tested separately, which ensures greater accuracy at each stage of the process. That way, you can be sure each piece is working properly before putting everything into production. ✍️
How Acumatica protects sensitive data
Data protection is usually the first concern raised when AI enters the picture, and that makes total sense. Acumatica tackles this challenge head-on by giving users control over which data should be masked before it leaves the system.
Sensitive parameters are:
- Encrypted and masked within the database
- Masked before data is sent to the third-party LLM
- Unmasked only after the response returns to Acumatica
The masking flow works like this, step by step:
- Data is masked inside Acumatica
- Masked data is processed by the LLM
- The response returns with masked values
- A retrieval function unmasks the data before users see it
Based on the process documented by Acumatica, sensitive data is masked before transmission and restored only within the application environment. At no point does unmasked sensitive data exist outside the system. That is a massive reassurance for anyone working with confidential information. Your data never leaves exposed to the outside world.
While many fields can be masked, there are some limitations. For instance, portions of larger free-text fields cannot be selectively masked. Understanding these nuances from the start helps teams design safer prompts. That is the kind of knowledge that prevents unpleasant surprises later on. 🔐
Centralized system instructions for long-term consistency
Instead of recreating the same rules over and over (which is a huge waste of time), Acumatica allows organizations to define reusable system instructions in a centralized library. This saves effort and ensures everyone is on the same page.
These instructions can include:
- Safety guidelines to prevent harmful outputs
- Security standards related to access and confidentiality
- General communication principles and response formats
Prompt engineers can then adapt these instructions as needed, ensuring consistency across different departments and use cases. And there is more: this approach reduces dependence on so-called tribal knowledge — that expertise that lives only in a few people’s heads and gets lost when they leave the company. With a centralized library, knowledge stays documented and available to everyone.
The real impact for teams and businesses
From a practical standpoint, what Acumatica‘s AI automation framework delivers is a significant reduction in time spent on repetitive, low-value tasks. Processes like document classification, field population, or summary creation can be automated with high accuracy, freeing teams to focus on more strategic analysis and decisions that truly require human judgment.
For IT and security leadership, the framework represents a major shift in how AI is perceived internally. Instead of being seen as a black box that nobody controls, AI automation becomes a manageable asset with clear policies, auditable records, and well-defined responsibilities. This makes conversations with legal, compliance, and audit teams much easier — teams that are often the most cautious when it comes to adopting new technologies, and rightfully so, since they are the ones responsible for ensuring the organization operates within legal and regulatory boundaries.
The launch of the AI Automation framework in Acumatica‘s 2026 R1 version is a clear signal that the ERP market is maturing in its relationship with artificial intelligence. It is no longer about adding AI as a marketing differentiator, but about integrating it in a structured way, with real governance, effective data security, and support for multiple language models. This combination creates the ideal conditions for companies of different sizes and industries to adopt automation with confidence, knowing they are in control of the technology — and not the other way around. ✅
