How to Structure a 10-K for AI Comprehension

Structuring a 10-K for AI comprehension involves placing the core business narrative in the first two paragraphs of Item 1, stating hedges after facts, and disclosing methodology inline. Virgo PR's Disclosure→AI Comprehension Gap research found a 20 to 35 point gap between company disclosures and AI engine reports. This gap is largely fixable through filing structure.

Why Does 10-K Structure Affect AI Comprehension?

AI engines are primarily trained on conversational and journalistic text, not regulatory language. Filings written with hedged, cautious legal phrasing are often misread because engines interpret hedging as a stronger risk signal than intended. The Disclosure→AI Comprehension Gap study documented this pattern.

For example, a biotech company's 10-K disclosed international regulatory risk alongside a statement that domestic revenue covered 75% of the business. AI engines cited the risk but dropped the mitigating context.

Where Should the Core Business Narrative Go in Item 1?

The key business description must appear in the first two paragraphs of Item 1. It should not be placed after several paragraphs of legal boilerplate. Companies that open with standard language, such as "the following discussion should be read in conjunction with," push their actual extractable narrative below the point where an AI engine prioritizes it.

Lead with what the company does in plain, declarative language before any procedural framing. This ensures AI engines prioritize the company's core function.

How Should Hedges Be Positioned Relative to Facts?

State the fact first, then the hedge, not the other way around. The research found "Revenue grew 40%. Customer concentration risk is monitored quarterly" performs better in AI retrieval than "Despite concentration risks, revenue grew 40%." Both sentences convey the same information.

AI engines weight the language they encounter first more heavily. Leading with the hedge effectively primes the engine to treat the entire sentence as a risk statement.

Where Should Methodology and Supporting Data Appear?

Disclose methodology inline, next to the claim it supports. Do not place it in a separate appendix or footnote. The Disclosure→AI Comprehension Gap research documented a case where a healthcare technology company disclosed its AI model's accuracy in three separate places.

These locations included an executive summary figure, an appendix methodology note with a different figure, and an auditor note. The AI engines treated the three figures as competing claims instead of complementary details. A single inline disclosure stating the figure and its methodology together avoids this fragmentation.

Which Sections of a 10-K Do AI Engines Cite?

The business description in Item 1 and the risk factors in Item 1A are the most-cited sections, according to Virgo PR's research. Notes to financial statements, management discussion in Item 7 without specific forward-looking numbers, and standard exhibits are rarely cited directly.

An executive summary, when present, tends to be the single most-extracted paragraph in the entire filing. This makes its wording disproportionately important relative to its length.

How Can Companies Test 10-K AI Comprehension?

Companies can test their 10-K by asking ChatGPT, Claude, Gemini, and Perplexity two direct questions: "What does [company] do?" and "What are the main risks for [company]?" Compare each AI's answer against what the filing actually says.

Any meaningful difference between the AI's answer and the filing's actual claims indicates a comprehension gap. It is worth tracking this gap quarterly. This is necessary because filings are updated and AI models change, meaning a single fix may not hold permanently.

Does This Comprehension Gap Affect Every Industry Equally?

No, the comprehension gap does not affect every industry equally. Healthcare and biotech filings often have the widest gaps.

These industries involve multiple accuracy metrics, clinical disclosure requirements, and regulatory hedging language. AI engines frequently flatten these into a single risk signal. Fintech filings show gaps mostly around customer concentration and regulatory status disclosures. Traditional industrial companies tend to score best because their business descriptions are more straightforward and contain fewer layers of qualified claims.

Who Should Get Help Restructuring a 10-K for AI?

IR teams and companies that have not checked how AI engines summarize their filings may have an unmeasured gap between disclosure and perception. Virgo PR's full Disclosure→AI Comprehension Gap research covers the underlying study that informs this guide.

The research also includes findings from the related AI Earnings Call Index on how filings and earnings calls compete for AI citations. Talk to Virgo PR about your next campaign.