​The AI Advantage Hiding In The Foundation Of Your Drupal CMS

Jon Stewart is the President of ZenSource, an open source platform combining CMS, enterprise level cloud hosting and support.

gettyThrough my work helping organizations build and manage Drupal websites, I know that Drupal has never been the loudest name in content management. While newer platforms often competed on simplicity and flash, Drupal kept investing in something less glamorous: structure, governance and clean data. Now, as AI changes how people find information, those quieter investments are paying off.

Search is shifting from ranking pages to being understood. AI assistants and answer engines don’t just index your site; they also extract, summarize and cite it. A buyer researching vendors increasingly gets an AI-generated answer before visiting a single website. Whether your organization appears in that answer depends less on keywords and more on how clearly your content can be interpreted. Platforms built on structured, well-governed content are the ones best suited to help companies succeed.

If you’ve been on Drupal for years, you’ve likely been doing AI optimization without calling it that.

Semantic URLs: Clean, human-readable paths signal topic and hierarchy to crawlers and language models, telling an AI system what a page is about before it reads a word.

Structured Content: Content types, fields and taxonomies break information into labeled pieces. An event has a date, a location and a speaker. That is exactly what retrieval systems need to answer questions accurately instead of guessing from a wall of text.

Machine-Readable Metadata: Schema.org markup, metatags and sitemaps help AI identify entities, authors and relationships.

API-First Architecture: Content flows through standard APIs to AI applications, agents and new channels without rebuilding.

Accessibility And Multilingual Support: Semantic markup built for screen readers is also easier for machines to parse, and structured translations extend your visibility across markets.

Governance: Revisions, workflows and permissions keep content accurate. When AI quotes you, accuracy is your brand.​​

​When I help long-time Drupal organizations prepare for AI-driven discovery, there are two key points I focus on: making structured content easy to maintain and introducing AI capabilities without sacrificing control. ​​

In my work, one of the most common reasons content models break down is that editors bypass structure when it slows them down. In higher education, for example, I regularly see program pages rebuilt as free-form layouts, with deadlines and tuition tucked into tables or images rather than maintained in dedicated fields. The information is available to human readers, but it becomes harder for AI systems to retrieve and interpret reliably.

This is why the authoring experience matters as much as the underlying architecture. For Drupal users, visual editing tools like Drupal Canvas can help bridge the gap between editorial flexibility and structured data, but it’s also important to design content templates that give editors flexibility while keeping key information in dedicated, reusable fields.

AI can help organizations retrieve information, answer questions and identify opportunities to improve existing content. But the ability to generate a plausible response does not establish that the response is accurate, current or appropriate to publish.​

In practice, this means grounding AI applications in trusted organizational content, making their sources visible and establishing clear approval requirements for consequential changes. It also means considering how easily an organization can change AI providers as capabilities, costs and requirements evolve. An architecture that separates content management from model selection gives technology leaders more flexibility without requiring them to surrender governance.​​

These principles can ultimately apply regardless of which content management system an organization uses, and the objective is to make reliable information easier for both people and machines to use while preserving the controls that keep it trustworthy.​​​​

• Audit your content model. Years of quick fixes often leave important information buried in body text. I regularly find event dates, locations and eligibility rules written into paragraphs, which AI assistants then misread or skip. Usually the fields to hold that information already exist and simply went unused, so the fix is migration and editorial discipline, not a rebuild. Moving key details into dedicated fields makes them visible to AI.

• Expand structured data. Make sure organizations, people, events, articles and FAQs carry consistent Schema.org markup.

• Clean up URLs and redirects. Consistent alias patterns and resolved redirect chains help crawlers and AI systems map your site.

• Retire stale content. Archive or update outdated pages so AI isn’t summarizing information you no longer stand behind.

• Add an accuracy check to governance. Most workflows ask whether content is ready to publish. For high-stakes content such as pricing, eligibility and policy, also ask whether you’d stand behind it if AI quoted it out of context.

• Stay current. Running a supported version is the gateway to the AI capabilities arriving now.

• Is content structured or trapped in page-sized rich text fields?

• Can AI tools and agents access it through open APIs and standards?

• Can you switch AI models as the market changes?

• Do humans stay in control, with review, audit trails and permissions?

• Do you own your data and hosting choices?

A “no” to any of these points is a sign that it might be worth further developing your strategy.

Many long-time Drupal organizations are approaching a major version upgrade, a consolidation or a redesign. Treat it as a content modeling exercise, not a lift-and-shift. Moving messy content to a newer version simply relocates the mess. The AI value comes from restructuring as you go: defining content types around how your audiences ask questions, establishing consistent taxonomies and setting governance rules from day one.

AI is also changing the economics of this work. It can help map legacy content to new structures, flag duplicates, suggest tags and clean up inconsistencies that once required extensive manual effort. Discovery, stakeholder review and QA still take real time, but the heavy lifting in between is faster than many organizations expect.

The organizations that win in AI-driven discovery won’t be the ones with the most AI features. They’ll be the ones whose content is easiest to understand, trust and reuse.​

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?