Plans to Mark AI Content Stir Controversy

A new proposal has been put forward to mark specific sections of web pages that are AI generated, using existing semantic HTML, in order to meet upcoming EU regulatory requirements. The idea has sparked debate among web developers and accessibility experts, with some questioning whether it is the most appropriate approach.

The proposal, created by David E. Weekly, aims to introduce an HTML attribute that signals to crawlers which parts of a page are generated by AI. While other proposals focus on identifying entire web pages as AI authored, Weekly’s approach targets individual sections within a page that otherwise contains human-written content.

Many modern web pages contain a mix of human-written and AI-generated material. News websites, for instance, might include a sidebar summarising articles using AI while keeping the main reporting written by humans. This partially AI-generated content is exactly what Weekly’s proposal seeks to address.

The proposal suggests using the <aside> HTML element, a core component of Semantic HTML, in combination with a new attribute to mark AI-generated sections. The <aside> element is traditionally used for content that is tangential or supplementary, such as sidebars, related links, or call-out boxes.

Weekly explains that existing methods for AI disclosure, such as page-level <meta> tags or HTTP response headers, cannot identify individual sections of content. “A news article page might contain a human-written investigation alongside an AI-generated summary sidebar,” he notes. “Current approaches only support page-level disclosure or response-level signals, which leaves a gap for marking individual sections.”

The regulatory context is clear: Article 50 of the EU AI Act, which comes into effect in August 2026, requires machine-readable labelling of AI-generated text. This creates a legal need for standards capable of marking specific sections rather than entire pages.

However, the proposal has generated discussion around whether <aside> is the correct semantic choice for AI-generated summaries. By definition, <aside> is intended for content that is indirectly related to the main document. Summaries, on the other hand, condense the core content and are therefore directly related. Critics argue that using <aside> may not be semantically correct in this context.

The HTML specification defines <aside> as: “The <aside> element represents a portion of a document whose content is only indirectly related to the document’s main content. Asides are frequently presented as sidebars or call-out boxes.” This highlights a potential tension between regulatory compliance and semantic accuracy.

In practical terms, the debate centres on whether AI-generated summaries that are closely tied to the main content should be treated as tangential. Some developers argue that the solution prioritises compliance over improving web usability or accessibility. Others see it as a pragmatic way to meet legal requirements using existing HTML structures.

The GitHub discussion surrounding the proposal reflects these concerns. Many contributors question the real-world benefits of adding section-level markup solely for legal compliance, noting that it may not provide meaningful advantages to users or search engines beyond satisfying the regulation.

One commenter summarised the concern: “The approach seems primarily aimed at meeting regulatory requirements, without a clearly demonstrated benefit for the web ecosystem as a whole.” This highlights a broader tension between law-driven technical standards and best practices for web design and accessibility.

Accessibility experts have also weighed in. Using <aside> for content that is semantically part of the main article could create confusion for assistive technologies, which rely on accurate HTML semantics to present content to users effectively. Misusing the element could unintentionally reduce accessibility.

Despite these criticisms, the proposal continues to be actively discussed and refined. Supporters argue that the new attribute offers a practical solution to an emerging regulatory challenge and provides publishers with a standardised way to disclose AI-generated sections.

The key takeaway is that the proposal is not yet settled. While it addresses an urgent compliance need, questions remain about whether it fully aligns with semantic HTML principles and best practices for content accessibility.

Ultimately, this debate reflects the broader challenges of integrating AI into web publishing. Publishers, developers, and regulators must balance transparency, usability, accessibility, and legal compliance as AI-generated content becomes increasingly common across news sites, blogs, and other digital platforms.