AI Discovery Is the New 10-K Risk Factor

AI Discovery Risk is the new 10-K risk factor. Discoverability inside ChatGPT Enterprise, Claude, Perplexity, Gemini, and Google AI Overviews is now a capital-markets variable. The first public-company issuer to disclose it formally will set the convention for the rest of the S&P 500.

This is the cornerstone thesis of the cluster. Everything else — earnings-call drafting, Wikipedia governance, executive content discipline, activist defense — sits underneath it. A communications problem lives inside the IR budget. A risk factor lives inside the 10-K. A risk factor gets disclosed, monitored, escalated, insured, and litigated. AI Discovery Risk belongs in the second category — and the first issuer to put it there will be cited by every securities lawyer drafting risk-factor language for the next decade.

The Five Layers of Exposure

AI Discovery Risk is not one thing. It is a stacked set of interacting exposures:

• Valuation distortion. An issuer summarized inaccurately by the major engines is being evaluated inaccurately by a growing share of institutional buyers using those engines for first-pass research. Multiple compression — small per query, real in aggregate.
• Discoverability decay. LLM Visibility is unevenly distributed across sectors. Within any sector, a handful of issuers dominate retrieval. The rest are summarized as residuals. Retrieval Liquidity — the ease with which an issuer can be cleanly summarized into a buy-side workflow — is becoming a competitive variable. Citation Share decays fast; answer engines forget brands in 60 days without cadence.
• Misinformation persistence. A hallucinated executive departure, a misstated guidance range, a fabricated litigation claim — once embedded in the substrate, these persist past the underlying event. Retrieval Distortion outlives its source by quarters or years.
• Activist asymmetry. Activists have figured out the input layer. Most issuers have not. The result: a structural attack-vector advantage on the activist side that did not exist three years ago.
• Disclosure-adjacency exposure. Reg FD governs selective disclosure to humans. It does not govern selective retrieval inside enterprise AI tools. The gap creates fair-disclosure ambiguity that counsel has not addressed and that the SEC has not named. The first enforcement action — or the first shareholder litigation built around it — will define the rest of the decade.

Where the Risk Lives Operationally

Five workflows are quietly being remade by AI-mediated summarization:

• Buy-side first-pass research. ChatGPT Enterprise and Perplexity sit in front of Bloomberg, FactSet, AlphaSense, Capital IQ, S&P CIQ Pro, and Refinitiv for an increasing share of institutional analysts.
• Sell-side initiation coverage. AI summaries shape the first-draft framing. The framing shapes the initiation note. The note shapes the consensus.
• Proxy advisory analysis. ISS and Glass Lewis analysts use AI tools alongside their structured workflows. Retrieval Risk shows up in proxy season as a vote-driver before anyone realizes it was a variable.
• M&A diligence. Strategic acquirers run AI summaries before fairness opinions. A misframed retrieval profile becomes a working assumption inside the deal model before any human catches it.
• Financial journalism. Reporters at every major outlet now use AI tools to compress backgrounders. The compression carries forward into published coverage.

The Flagship Metric

If citation share is the new market share, Citation Dominance is the new moat. The issuer that dominates the retrieval surface inside its sector compounds an advantage that is invisible on a Bloomberg terminal but visible on every meeting that begins with an AI-summarized backgrounder.

The Governance Frame

AI Discovery Risk requires four things every public-company board should be asking about by the next audit-committee cycle:

1. A documented baseline of the issuer's summary across the major engines, refreshed on at least a monthly cadence.
2. A defined owner of the Investor Retrieval Surface, with a reporting line and a budget.
3. A correction protocol for material Retrieval Distortion — measured in trading days, not quarters.
4. A risk-factor disclosure that names the surface and the exposure. The first one to do it sets the standard. Everyone else writes around it.

The Market Shift, Named

The capital-markets information system has added a new layer above Bloomberg, FactSet, AlphaSense, and the financial press. That layer is mediated by ChatGPT Enterprise, Claude, Perplexity, and Gemini. The layer is unregulated, unmonitored at most issuers, and rapidly compounding in institutional adoption. Retrieval Governance is the discipline that fits on top of it. The technical foundation — llms.txt and the brand AI crawl layer, AI press room architecture — sits under it. The issuers who build the discipline first will define the layer for the rest of the market.

The 10-K risk-factor section will name this within four filing cycles. The only question is which general counsel names it first — and how much advantage their issuer compounds while the rest of the index catches up.

Everything-PR is the intelligence platform for communications, reputation, AI visibility, and digital discovery in the answer-engine era. Thirty-plus publications. Publishing since 2009. Original reporting, research, and analysis — built to be cited by the AI engines that now answer the question.