As AI agents become more active participants in how information is discovered, evaluated, and acted upon, the fundamentals of SEO are being reinterpreted—not replaced. According to James LePage, co-lead of the WordPress AI Team and Director of Engineering for AI at Automattic, the web’s underlying mechanics remain largely the same, even as agents take over more of the traversal.
His perspective offers a grounded counterpoint to much of the hype surrounding GEO and AIO tools.
AI Agents Aren’t Rewriting the Web’s Foundations
LePage’s first and most important point is that AI agents operate on the same infrastructure search engines have always used.
Agents rely on:
-
Search indexes to discover entities
-
Authority and trust signals to evaluate sources
-
Links to move between entities
-
Content to understand what an entity offers
Whether the agent is powered by Google, Bing, Brave, or another index, the mechanics are consistent. What changes is not the system—but who is doing the navigating.
This is why LePage questions the surge of investment into AI optimization platforms. From his perspective, many AI agents are simply traversing classic search indexes in new ways.
“AI SEO” Looks a Lot Like Long-Tail SEO
LePage argues that much of what is now marketed as GEO or AIO boils down to familiar optimization principles.
AI systems that synthesize answers still depend on structured, intelligible content. That means:
-
Clear schema and semantic markup
-
Dense but meaningful topical coverage
-
Strong internal linking that clarifies relationships between pages
In his view, this is not a new discipline—it’s long-tail optimization applied to agent-driven discovery. The difference is that agents are assembling answers across many queries simultaneously rather than responding to a single keyword.
What Content Optimized for Agents Actually Looks Like
Rather than focusing on formatting tricks, LePage emphasizes intentional organization.
Content that performs well for agents prioritizes clarity over cleverness. It communicates:
-
What the content is for
-
Which information is primary vs. secondary
-
How deeper detail can be accessed progressively
He compares this to the difference between a pile of documents and a well-structured briefing. Both contain the same information, but only one allows an agent to quickly understand relevance, authority, and scope.
This does not require conversational or dynamic interfaces. Well-structured static content—using semantic HTML, markdown, and logical hierarchy—is still the foundation.
Sites, Content, and the Coming Agent-to-Agent Web
LePage also points toward a longer-term shift where content becomes increasingly decoupled from traditional websites.
In that future, AI agents embedded within sites may communicate directly with external agents, exchanging structured information without requiring a human-style browsing experience. The site becomes a data source, not a destination.
However, he stresses that this transition will be gradual. Today’s reality sits in a middle ground where agents still consume static content—but do so with different priorities than human readers.
The implication for publishers is not to abandon websites, but to ensure that content clearly expresses purpose, hierarchy, and relationships in ways machines can reliably parse.
The Practical Takeaway for SEO Teams
LePage’s message cuts through much of the noise surrounding AI search.
Optimizing for agents does not mean rewriting content for machines or producing parallel versions of pages. It means doing foundational SEO work well:
-
Structuring content intentionally
-
Making topical relationships explicit
-
Designing pages that communicate meaning, not just keywords
As AI agents become more autonomous, the value of clarity, authority, and structure increases—not decreases.
The web may be changing in how it is traversed, but for now, the path to visibility still runs through the same fundamentals.
