When it comes to writing Standard Operating Procedures (SOPs), AI has become the technical writer’s best friend.
However, human interaction—aka, the technical writer—is still needed in order to steer this marvel of algorithm genius down the right road.
That’s particularly important when it comes to creating SOPs for today’s highly regulated utilities, an industry that is increasingly recognizing the value of AI, but cautiously so.
And for good reason.
At its core, AI uses probabilistic thinking. Utility SOP documentation requires deterministic thinking, which breeds certainty. And no industry relies more on certainty than America’s utilities, where the slightest misstep can be a monumental issue.
While it brings much to the table, AI on its own can’t achieve this. It needs technical writers to ensure output is accurate to a specific utility in a specific jurisdiction regarding very specific procedures with full compliance.
We’ll examine that here, including where AI falls short, and how the technology can be most useful to utility SOP development.
Unpredictability: The New Challenges
Utility usage was once fairly predictable, especially where electricity was concerned. One centralized company delivered wired power. The meter got read, and the bill went out. Customers had few options, and only complained when the lights wouldn’t come on.
Unlike today, we didn’t have data centers requiring massive power and water usage, or EV charging stations, potentially challenging the grid. There wasn’t such a demand for cleaner, more reliable energy alternatives or choices in providers. And the type of economic and technological transitions we experience nowadays is a comparative drop in the bucket.
Suffice it to say, utilities deal with far more unpredictability than they used to.
Building SOP Predictability: Why Technical Writers Matter
The purpose of any SOP is to build procedural, regulatory, and operational predictability. It’s about managing the inevitable and being ready when it happens.
Simply put, a well-written, enforced SOP means that everyone is on the exact same page doing things the same way every time—without compromise, uncertainty, or risk of misinterpretation. It makes compliance easier to demonstrate to regulators, and can even save lives.
That well-written utility SOP starts with AI, and ends with a knowledgeable technical writer.
Having AI on board adds speed, efficiency, structure, and breakthrough technology to the SOP process. Having a technical writer on board adds contextual understanding through validation and verification of what AI generates.
Where AI Falls Short
Let’s look at a few examples of where AI definitely requires human technical writer involvement.
Prone to Inaccuracies
On its own, AI can generate well-written, seemingly plausible, yet incorrect output. Without the technical writer’s supervision, the wrong procedural information can lead to safety and compliance issues.
Lacks Domain-Specific Judgment
AI doesn’t truly understand utility operations, hazards, or equipment. Nor does it have any local context to go on. Without an SME review, critical details and jurisdiction-specific rules can be missed.
Doesn’t Fully Understand Regulatory Standards
AI relies on generalized language. It doesn’t know specific regulatory requirements like those for OSHA, EPA, PHMA, or NERC.
Lacks Controlled Document Governance
AI doesn’t create audit trails, version controls, or approval workflows by itself. Utility regulatory compliance depends on controlled document governance, which AI tools don’t inherently provide.
Lacks Human-Factor Insight
AI can write text, but doesn’t design for real-world use cases—such as usability during emergencies, stress conditions, or field validation contexts where clarity and cognitive load matter.
Potential Security and Data Risks
Public AI platforms might store or use proprietary SOP content in ways that jeopardize security or confidentiality—a significant concern for utilities managing critical infrastructure.
Understandably, these shortcomings can give anyone extra pause. But according to an article written by Aaron Gofeder, CEO of AZX, which develops AI solutions for energy companies, there is much good news.
“Utilities sit on mountains of unstructured data, including spreadsheets, PDFs, microfiche, and handwritten notes. Unlocking that data used to take armies of consultants, or it would just sit unused, orphaned off in a PDF black hole. Now, a well-tuned AI agent can handle it quickly and turn it into value.”
Gofeder also notes, “You don’t have to believe in artificial general intelligence for this. Instead, technology can just be a better way to read what we already have.” Gofeder is also quick to support human-AI collaboration as critical for ensuring clarity and accuracy.
Where AI is Useful
AI-generated documentation fundamentally cuts down on a huge amount of manual work that technical writers typically must do. That includes the laborious task of drafting from scratch, as well as the need to cross-reference, standardize, and search across a vast sea of data.
Faster Drafting
AI generates structured SOP drafts, allowing the writer to focus on validation and accuracy, saving time in the process. This is valuable for utilities due to the often large libraries of procedures they must maintain.
Ensures Consistent Structure and Terminology
AI excels at applying consistent formatting and terminology. It can do this across multiple SOPs and standardize documentation across departments and regions, thereby reducing field crew misinterpretation.
Assists with Large-Scale Updates
In safety-critical environments, regulatory requirements change, and procedures evolve frequently. When this occurs, AI can quickly revise and regenerate sections of relevance, saving significant time.
Support Exploratory and Research Tasks
When mapping regulatory requirements in SOPs, AI reduces how long it takes SMEs to gather regulatory context. AI can quickly survey regulatory guidance, summarize standards, or identify relevant sections of complex rules.
Utilities Need AI and Technical Writers to Think for Each Other
AI adds an unprecedented level of structure, consistency, uniformity, and speed to nearly every aspect of the SOP process.
On its own, however, AI output is only as whiz-bang as what technical writers prompt it to put out.
That output is especially crucial to utilities, where an SOP can be the difference between customer comfort and complaint, regulation, and fine. Even life or death.
Technical writers can gut-check and validate every prompt with proprietary awareness and wisdom. It’s fair to say that AI is there to do the dirty work, and technical writers are there to tidy up.
AI and technical writers need each other. And utilities need both now more than ever.
