Agentic AI and Brand Differentiation Risks

James LePage, Director of Engineering for AI at Automattic and co-lead of the WordPress AI Team, has outlined a potential shift in how websites may operate in the coming years. According to LePage, the rise of agentic AI could transform sites into interactive data sources, potentially diminishing the unique branding and voice that sites currently offer.

LePage describes a future where AI agents interact with websites autonomously, performing tasks at a granular level while humans operate at a higher strategic level. In this model, site owners set guidelines and review outcomes rather than approving every individual action.

He outlines three stages of AI autonomy. The first is similar to today’s advanced web search: AI gathers content, synthesises it, and presents it, but users still take the main actions. The next stage allows agents to carry out specific tasks like purchases or bookings within clearly defined boundaries. Eventually, agents could operate semi-independently, following standing guidelines and acting almost as economic actors in their own right.

One of the key implications LePage highlights is that websites may increasingly function as data sources rather than immersive experiences. AI agents visiting a site extract the information they need and present it according to their own logic, rather than the way a human designer or content creator intended. This could flatten the perceived differentiation between sites.

LePage warns that this shift has particularly complex consequences for informational and service-oriented websites. While commercial sites may benefit from AI-driven efficiency in transactions, media and informational sites could see their brand, voice, and perspective diluted when content is summarised alongside competitors.

To address this, LePage suggests leveraging AI on the website itself. By deploying site-specific AI agents, websites can provide interactive and personalised experiences that engage users beyond the static presentation of content. These agents could communicate with visiting AI agents, ensuring that information is represented accurately and effectively.

This agent-to-agent interaction model allows a site to maintain influence over how its content is interpreted, potentially preserving its unique value and branding in an AI-driven ecosystem. The goal is to create experiences that are personalised, immersive, and capable of handling complex user interactions without requiring manual intervention for each variation.

LePage emphasises that the rise of agentic AI does not eliminate the need for human oversight but shifts the nature of that oversight. Humans will guide strategy and goals, while AI handles detailed execution, content synthesis, and user interaction.

This approach could redefine website design and the online user experience. For example, sites might include interactive data visualisations, product configurators, or personalised content flows, raising the standard for what constitutes a valuable visit. Information alone will no longer be the key differentiator—experiential and interactive elements will take centre stage.

For informational sites, however, the challenge is more pronounced. If AI agents can extract and deliver content elsewhere, the intrinsic value of the website itself diminishes. Without adaptation, these sites risk becoming mere data points to be scraped rather than active participants in user interactions.

LePage suggests that the evolution of the web may see sites and users represented by their own AI agents. These agents would communicate and negotiate on behalf of their principals, exchanging information, applying constraints, and reflecting preferences in real time. The infrastructure for this agent-to-agent web is already under development, with contributions from major players including OpenAI, Google, Microsoft, and Anthropic.

He warns that entities that fail to adopt this model will fall behind. Their content may still exist online, but it will be passively consumed rather than actively engaged with. Websites that integrate AI effectively will be able to participate in richer, more controlled exchanges, preserving brand influence and user experience.

Ultimately, LePage sees this evolution as both a challenge and an opportunity. Publishers and service providers who embrace AI-driven interactions may find new ways to differentiate themselves, while those who resist risk losing their distinct voice in a rapidly changing digital landscape.

In conclusion, the emergence of agentic AI may flatten traditional branding and voice, but it also opens doors for innovative ways to engage audiences. The future of the web may not be static content alone—it will be defined by dynamic interactions between AI agents representing both sites and users.