Not long ago, investors doing preliminary research on a public company would type a name into a search engine, scan a page of blue links, and decide what to click. That experience is changing. For a growing share of users, the first result is no longer a list of links—it is a short paragraph generated by an AI system, synthesizing several sources into one answer. This shift has real implications for how public companies are perceived before an investor ever reaches a filing, a press release, or an earnings call.

Understanding it requires two terms: SEO and GEO. SEO—search engine optimization—is the established discipline of helping search engines find, index, and rank a company's content. GEO—generative engine optimization—is the newer practice of improving how a company is represented when AI systems such as ChatGPT, Google AI Overviews, Gemini, and Perplexity generate an answer rather than just a list of links. Put simply: SEO helps a company get found. GEO shapes what investors read first when they do.

The stakes are higher than they might initially appear. Investor perception does not begin at the moment someone reads a 10-K. It begins with the first impression—and increasingly, that impression is shaped by an AI summary.

The Scale of What Has Already Changed

The numbers behind this shift are significant. Alphabet reported on its Q1 2025 earnings call that Google's AI Overviews reached 1.5 billion monthly users, available in more than 200 countries. [1] OpenAI announced in February 2025 that ChatGPT had surpassed 400 million weekly active users, [2] a figure that climbed to over 800 million by October 2025. [3] Perplexity processed approximately 780 million search queries in May 2025 alone, according to CEO Aravind Srinivas at Bloomberg's Tech Summit 2025—up from 230 million the prior year. [4]

This volume is changing how information actually gets consumed. Bain & Company reported in February 2025 that about 80% of consumers rely on AI-generated summaries for at least 40% of their searches, and that roughly 60% of searches now end without the user clicking through to any destination website. [5] Organic web traffic to many sites is being reduced by an estimated 15% to 25% as a result. [6] Pew Research Center quantified the behavioral gap in July 2025: users who encountered an AI summary clicked on a traditional search result in just 8% of visits, versus 15% for searches without one—nearly half the click rate. [7]

For companies with strong public profiles, this is manageable. For smaller and mid-cap issuers with limited analyst coverage and a thinner web footprint, it is a material risk to how they are understood in the market.

SEO vs. GEO: A Working Distinction

SEO, in Google's own framing, is the process of helping search engines understand content and helping users decide whether to visit a site. In practice, this means clear page titles, logical site structure, crawlable links, and content that earns authority through relevance and helpfulness over time. [8] Google points to E-E-A-T—experience, expertise, authoritativeness, and trustworthiness—as the organizing principle for what ranks well.

For public companies, this has traditionally meant well-structured investor relations pages: stable URLs for SEC filings, clearly labeled earnings materials, governance documents that are easy to locate, and investor contact information that is not buried three clicks deep. Academic research has linked better IR disclosure to higher analyst following, more institutional investors, tighter bid-ask spreads, and lower information asymmetry in the market. [9] Good SEO is not a marketing exercise—it is part of disclosure hygiene.

GEO operates on the same public record but takes it further. Rather than ranking a company's pages, AI systems like ChatGPT Search, Perplexity, and Google AI Overviews synthesize information from multiple sources—filings, news, blogs, third-party databases, press releases—and compress them into a short narrative. ChatGPT Search describes itself as providing fast, timely answers with links to relevant sources. Perplexity searches the web in real time and distills it into a cited summary. Google AI Overviews generates an answer directly in the search results. [10] The key distinction: SEO presents choices; GEO presents a conclusion.

The practical difference is one of control. With SEO, a company controls its own pages, titles, and structure. With GEO, the company controls only the source material—not the wording or framing of the AI's response. Both Google and OpenAI openly acknowledge their systems can make mistakes. [11] A March 2025 Tow Center for Digital Journalism study, published in the Columbia Journalism Review and summarized by Nieman Lab, tested eight AI search engines and found citation errors in more than 60% of queries—with most systems defaulting to a confident wrong answer rather than declining to respond. [12] GEO is not about gaming the model; it is about improving the quality and consistency of the information the model draws from.

What Investors Now Encounter First

In the SEO era, an investor researching a company would typically see several results and choose which to click. In the GEO era, they may receive a synthesized paragraph before they ever reach an official source. If that paragraph is balanced and accurate, the company benefits from the efficiency. If it is stale, incomplete, or framed around an older negative event, the damage occurs before the investor reaches a primary source.

This is a particularly acute problem for certain types of content: litigation history, governance disputes, restatements, internal-control issues, or inconsistent financial disclosures. AI systems tend to surface what is widely documented and easily summarized—which often includes press coverage of adverse events more readily than subsequent remediation. A company that settled a lawsuit two years ago and meaningfully improved its controls may still receive an opening AI summary centered on litigation risk, simply because the older coverage is more indexed and more numerous than the later record.

The same risk applies to disclosure consistency. If a company describes a non-GAAP metric differently across a press release, its website, and earnings call commentary, an AI model may surface that inconsistency as a signal of credibility risk—not because the company intended it, but because the system is reconciling contradictory inputs. Narrative coherence is not just a communications preference; it is increasingly a factor in how the market first encounters a company's story.

Research supports the broader concern. A 2024 National Bureau of Economic Research working paper found that media-driven narratives significantly affect cross-border institutional investment flows, even after controlling for macroeconomic fundamentals, and that investors react more sharply to negative narratives than to positive ones. [13] AI answer engines compress a scattered narrative environment into one paragraph—which concentrates that dynamic.

Small and mid-cap issuers face a disproportionate version of this problem. SEC Commissioner Uyeda noted in April 2026 that 44% of small- and mid-cap stocks have no analyst coverage at all, limiting the depth of credible information available to the market. [14] The SEC's 2022 staff report on investment research for smaller issuers documented that reduced analyst coverage increases information asymmetry, reduces institutional presence, and can affect liquidity. [15] In a GEO environment, a thin analyst footprint means answer engines have fewer high-quality sources to work from—and are more likely to over-weight a stale news item or an outdated filing.

What Regulators Still Expect

The legal baseline for disclosure has not changed because the discovery layer has. Reg FD still requires that material information be broadly and non-exclusively distributed to the public. For non-intentional selective disclosure, the SEC requires prompt correction within 24 hours or by the next day's trading on the NYSE. The SEC has acknowledged that websites can be an important component of disclosure, but they are not automatically sufficient on their own. [16]

The SEC's March 2024 enforcement actions against investment advisers Delphia and Global Predictions—the agency's first AI-washing cases—show that exaggerated AI-related claims can create antifraud risk under existing federal securities law. Those matters involved registered advisers, not public issuers, but the underlying principle has broader resonance: the antifraud provisions apply to all material public-facing statements, and AI systems can amplify incomplete or inaccurate ones far beyond their original context. [17]

The SEC's plain-English rule from 1998 is also more relevant now than it may have been when first adopted. The Commission stated that clear, concise disclosure helps create a better informed securities market. [18] That principle has a new dimension: plain language is also easier for AI retrieval systems to process and summarize faithfully. Dense, passive-voice boilerplate is harder for models to interpret accurately, and harder for them to attribute correctly. [19] Inline XBRL tagging in filings, which the SEC describes as both human-readable and machine-readable, is part of the same logic—structured, labeled data travels more accurately through the information ecosystem.

Outside the United States, the EU AI Act introduces transparency obligations for certain AI systems, with the stated aim of reducing misinformation and helping people make informed decisions. [20] Those rules do not rewrite issuer disclosure law, but they point in the same direction: more transparency, greater traceability, and less tolerance for ambiguous public information.

CFA Institute reported in 2025 that AI and big data tools are seeing increasing adoption across investment management workflows. [21] FINRA's 2025 National Financial Capability Study found that 20% of U.S. adults expressed interest in receiving financial advice from AI—a signal that AI is already part of how information is gathered, triaged, and compared, even if formal investment decisions remain human-led. [22]

What Companies Can Do

The practical response to GEO risk is not technical sophistication. It is disciplined public communication applied consistently.

Consistency across primary sources. Make sure core facts—business description, key metrics, capital structure, strategy language—are stated consistently across filings, press releases, investor relations pages, earnings call materials, and governance documents. Inconsistency is the primary input that causes AI systems to summarize conflict where none was intended.

Machine readability. Keep IR pages structurally clean with stable URLs, descriptive headers, and no unnecessary crawler blocks. Use available tagging formats—Inline XBRL in filings, schema markup on web pages—to label financial data so it travels accurately through automated systems. Google's documentation and OpenAI's publisher guidance both point clearly in this direction.

Write for clarity, not for spin. Google explicitly warns against content produced to manipulate search rankings, and the SEC has long favored clear, concise disclosure over legalistic hedging. [23] In a GEO environment, vague or overqualified language can backfire—models often strip away the scaffolding and surface the underlying ambiguity.

Build an explanatory record. Publish authoritative content that addresses recurring investor questions: capital allocation priorities, how key non-GAAP metrics are calculated, risk factor context, governance structure, and the timeline of any material corporate events. This is especially important for smaller issuers with limited sell-side coverage. Better source material produces more accurate AI summaries.

Monitor what AI systems say about your company. Companies already monitor search results, media coverage, and data-vendor profiles. The same practice should extend to answer engines. When a summary is wrong, the long-term fix is to correct the underlying public record, date updates clearly, and strengthen the authority of the primary source.

The Bottom Line

SEO has not become obsolete. Discovery still precedes interpretation, and a company whose official content is hard to find, poorly labeled, or blocked from crawlers will underperform in both traditional search and AI-generated answers. The fundamentals—clear structure, helpful content, stable pages, authoritative links—remain the foundation.

What GEO adds is a new layer of investor-perception risk that sits on top of those fundamentals. As AI systems become a primary interface through which many investors first encounter a company, the quality of a company's public record is no longer just a compliance matter. It is a market perception matter. The companies that will navigate this environment best are those that treat disclosure quality not as a box to check—but as a competitive asset.

Sources

[1] Alphabet Q1 2025 Earnings Call (April 2025) — https://abc.xyz/investor/events/event-details/2025/2025-Q1-Earnings-Call/

[2] TechCrunch — OpenAI now serves 400 million users every week (February 2025) — https://techcrunch.com/2025/02/20/openai-now-serves-400-million-users-every-week/

[3] TechCrunch — Google's AI Overviews have 2B monthly users; ChatGPT context (July 2025) — https://techcrunch.com/2025/07/23/googles-ai-overviews-have-2b-monthly-users-ai-mode-100m-in-the-us-and-india/

[4] Backlinko — Perplexity AI Statistics (citing Bloomberg Tech Summit, May 2025) — https://backlinko.com/perplexity-statistics

[5] Bain & Company — Consumer Reliance on AI Search Results (February 2025) — https://www.bain.com/about/media-center/press-releases/20252/consumer-reliance-on-ai-search-results-signals-new-era-of-marketing--bain--company-about-80-of-search-users-rely-on-ai-summaries-at-least-40-of-the-time-on-traditional-search-engines-about-60-of-searches-now-end-without-the-user-progressing-to-a/

[6] Bain & Company — Goodbye Clicks, Hello AI (2025) — https://www.bain.com/insights/goodbye-clicks-hello-ai-zero-click-search-redefines-marketing/

[7] Pew Research Center — Google Users Are Less Likely to Click on Links When an AI Summary Appears (July 2025) — 8% click rate with AI summary vs. 15% without — https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/

[8] Google Search — SEO Starter Guide — https://developers.google.com/search/docs/fundamentals/seo-starter-guide

[9] Chang et al. (2008) — Internet-Based Corporate Disclosure and Market Value, SAGE Journals — https://journals.sagepub.com/doi/10.1177/031289620803300208

[10] OpenAI — Introducing ChatGPT Search — https://openai.com/index/introducing-chatgpt-search/

[11] Google — AI Overviews Help & Accuracy Guidance — https://support.google.com/websearch/answer/14901683?hl=en

[12] Tow Center for Digital Journalism / Nieman Lab — AI search engines fail to produce accurate citations in over 60% of tests (March 2025); full study at Columbia Journalism Review — https://www.niemanlab.org/2025/03/ai-search-engines-fail-to-produce-accurate-citations-in-over-60-of-tests-according-to-new-tow-center-study/

[13] NBER Working Paper w33159 — Media Narratives and Cross-Border Investment (2024) — https://www.nber.org/papers/w33159

[14] SEC Commissioner Uyeda — Remarks at SBCFAC (April 28, 2026) — https://www.sec.gov/newsroom/speeches-statements/uyeda-remarks-sbcfac-042826

[15] SEC Staff Report — Investment Research for Small Issuers (2022) — https://www.sec.gov/files/staff-report-investment-research-small-issuers.pdf

[16] SEC — Regulation FD Final Rule (2000) — https://www.sec.gov/rules-regulations/2000/08/selective-disclosure-insider-trading

[17] SEC — Press Release No. 2024-36: Charges Against Delphia and Global Predictions for AI-Washing (March 18, 2024) — https://www.sec.gov/newsroom/press-releases/2024-36

[18] SEC — Plain English Disclosure Rule (1998) — https://www.sec.gov/rules-regulations/1998/01/plain-english-disclosure

[19] SEC — Inline XBRL Structured Data — https://www.sec.gov/data-research/structured-data/inline-xbrl

[20] European Commission — EU AI Act Regulatory Framework — https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai

[21] CFA Institute — Creating Value from Big Data in Investment Management (2025) — https://rpc.cfainstitute.org/research/reports/2025/creating-value-from-big-data-in-the-investment-management-process

[22] FINRA Foundation — Sixth Wave National Financial Capability Study (2025) — https://www.finra.org/media-center/newsreleases/2025/finra-foundation-releases-sixth-wave-national-financial-capability

[23] Google Search Central — Google Search and AI Content (2023) — https://developers.google.com/search/blog/2023/02/google-search-and-ai-content