BlogFiling Analysis

AI Washing: What 10,000 SEC Filings Reveal About Corporate AI Hype

16 min readFiling Analysis

In 2023, only 12% of S&P 500 companies mentioned artificial intelligence in their 10-K filings. By 2025, that number exploded to 84%. But here is the uncomfortable question nobody on Wall Street wants to answer: how many of those companies are actually using AI, and how many are just stuffing buzzwords into their filings to ride the hype? The SEC is starting to ask the same question - and the answers could reshape how investors evaluate the AI trade.

The AI Disclosure Explosion

The surge: AI mentions in S&P 500 10-K filings went from 12% (2023) to 84% (2025)

The concern: Only 40% of S&P 500 provide substantive AI disclosures; 60% are generic

The crackdown: SEC has designated AI enforcement as a 2026 exam priority with active cases

The framework: Five-step method to distinguish real AI companies from AI washers

From 12% to 84%: The AI Disclosure Explosion

Two years ago, artificial intelligence was a niche mention in corporate filings - something tech companies discussed and everyone else ignored. Then ChatGPT went viral, and corporate America collectively discovered that "AI" was the most powerful two letters you could add to a 10-K filing.

The numbers tell a stunning story. According to research from The Conference Board and Harvard Law School Forum on Corporate Governance, AI disclosures in annual filings have undergone the fastest growth of any risk factor category in SEC history.

YearCompanies Mentioning AIStandalone AI Risk FactorAvg. AI Word CountTrend
202312%3%45 wordsBarely mentioned - AI was a niche topic
202465%14%280 wordsChatGPT effect - companies rush to mention AI
202584%36%520 wordsSEC scrutiny forces substantive disclosure

Why This Growth Rate Matters

For context, climate-related disclosures took nearly a decade to go from niche to mainstream in SEC filings (2012-2022). AI disclosures achieved the same penetration in just two years (2023-2025). This speed suggests that much of the growth is reactive - companies rushing to add AI language because their competitors did, not because they have genuine AI operations.

The Conference Board found that 72% of S&P 500 companies disclosed at least one material AI risk in 2025, up from just 12% in 2023. That is a 6x increase in two years - the fastest adoption of any risk factor category in SEC filing history.

Sector-by-Sector: Where AI Is Real and Where It Is Not

Not all AI disclosures are created equal. A technology company building neural networks has a fundamentally different relationship with AI than a cereal manufacturer claiming to use AI for supply chain optimization. Here is how the disclosures break down by sector:

Technology

98% disclosing
Avg Risk Factors:

4.2 AI-related risks

Primary Focus:

Competition, talent, IP protection

Assessment:

Mostly real - AI is core business

Financial Services

92% disclosing
Avg Risk Factors:

3.8 AI-related risks

Primary Focus:

Algorithmic bias, regulatory, cybersecurity

Assessment:

Mixed - some genuine, some compliance theater

Healthcare

87% disclosing
Avg Risk Factors:

3.1 AI-related risks

Primary Focus:

Diagnostic AI, FDA regulation, data privacy

Assessment:

Growing real use cases in diagnostics

Industrials

78% disclosing
Avg Risk Factors:

2.4 AI-related risks

Primary Focus:

Automation, predictive maintenance, supply chain

Assessment:

Often vague - many are aspirational

Consumer Staples

61% disclosing
Avg Risk Factors:

1.3 AI-related risks

Primary Focus:

Supply chain optimization, demand forecasting

Assessment:

Mostly buzzword stuffing

Utilities

45% disclosing
Avg Risk Factors:

0.9 AI-related risks

Primary Focus:

Grid optimization, predictive outage detection

Assessment:

Emerging but mostly aspirational

The Sharpest Growth: Financials and Healthcare

The sectors with the most dramatic increase in AI disclosures are financials (from 14 to 63 companies, +350%) and healthcare (from 5 to 47 companies, +840%). Financial companies face regulatory and reputational risks tied to algorithmic decision-making, while healthcare companies are navigating FDA requirements for AI-powered diagnostics. In both sectors, the regulatory pressure is forcing more substantive disclosures.

The Five AI Risks Corporate America Fears Most

What are companies actually worried about when it comes to AI? The risk factors tell us. Here are the five most commonly disclosed AI risks across the S&P 500, ranked by frequency:

#1: Reputational Risk

38% of S&P 500

Failed AI products, biased outputs, or AI-generated misinformation eroding brand trust

Example: 45 companies warned that AI projects failing to deliver could damage stakeholder confidence

Red Flag: Companies listing reputational risk without describing their AI products may be preemptively covering

#2: Cybersecurity Risk

20% of S&P 500

AI expanding attack surfaces while arming adversaries with more sophisticated tools

Example: Companies flagging that AI-powered phishing and deepfakes threaten internal security

Red Flag: This is a legitimate risk for any company, not just AI companies

#3: Legal and Regulatory Risk

18% of S&P 500

Uncertain regulatory landscape, IP ownership questions, liability for AI decisions

Example: Companies warning about potential SEC enforcement actions for misleading AI claims

Red Flag: Companies disclosing legal risk about AI they don't actually use

#4: Talent Competition Risk

15% of S&P 500

Inability to hire or retain AI engineers and researchers in a hypercompetitive market

Example: Companies disclosing that loss of key AI personnel could materially impact operations

Red Flag: If AI is in risk factors but not in CapEx or R&D spending, that's a signal

#5: Competitive Displacement Risk

12% of S&P 500

Competitors using AI more effectively could erode market position

Example: Companies warning that failure to adopt AI could result in loss of market share

Red Flag: The most common AI washing signal - fear of being left behind without concrete plans

The AI Washing Playbook: How to Spot Fake AI Companies

AI washing is the practice of companies misrepresenting or exaggerating their use of artificial intelligence to attract investors, boost stock prices, or appear competitive. The SEC has compared it to "greenwashing" in ESG - making claims that sound good but lack substance.

Here are the four most common patterns we see in SEC filings, ranging from blatant AI washing to genuine disclosure:

The Vague Promise

Likely AI washing

Company mentions AI repeatedly in 10-K but provides no specific use cases, revenue attribution, or CapEx allocation

Filing language: "We are leveraging artificial intelligence and machine learning across our operations to drive efficiency and innovation"

What to look for: Zero specifics on what AI does, how much it costs, or what revenue it generates

The Risk-Only Disclosure

Defensive posturing, not real AI adoption

AI appears only in risk factors section, not in business description or MD&A

Filing language: "The rapid development of AI technologies could disrupt our industry and business model"

What to look for: If AI is only a threat and never an opportunity, the company probably is not using it

The CapEx Mismatch

Claims without investment = AI washing

Company claims AI transformation but capital expenditure and R&D spending are flat or declining

Filing language: "We are making significant investments in AI capabilities across our platform"

What to look for: Compare CapEx and R&D trends year-over-year. Real AI investment requires real spending

The Genuine Disclosure

Real AI adoption

Company provides specific AI use cases, quantified metrics, dedicated CapEx, and measurable outcomes

Filing language: "Our AI-powered recommendation engine contributed $2.3B in incremental revenue, representing 12% of total sales"

What to look for: Specific dollar amounts, percentages, named products, measurable KPIs

The SEC Crackdown: AI Washing Is Now an Enforcement Priority

The SEC is not just watching the AI disclosure explosion - they are actively cracking down on companies that overstate their AI capabilities. Here is the enforcement landscape as of early 2026:

SEC Investor Advisory Committee

December 2025

Voted to recommend mandatory AI disclosure guidelines for public companies

Called for issuers to define AI usage, disclose board oversight mechanisms, and report on material AI deployments

Status: Recommendation phase - not yet a rule

SEC Division of Enforcement

2024-2025

Multiple enforcement actions against companies making misleading AI claims to investors

Companies faced penalties for exaggerating AI capabilities in marketing materials used for fundraising

Status: Active enforcement

DOJ + SEC Joint Actions

2025

Criminal and civil cases against companies misrepresenting AI technology to investors

DOJ pursuing wire fraud charges alongside SEC civil enforcement for AI washing in investment pitches

Status: Precedent-setting cases underway

SEC 2026 Exam Priorities

November 2025

AI and emerging technology designated as key examination focus area

Examiners will evaluate whether AI-related representations are accurate and whether technology-driven recommendations align with regulations

Status: Active - exams beginning 2026

The Board Oversight Paradox

Here is a fascinating finding: while public AI disclosures are declining in specificity (companies are getting more cautious about making claims), board-level AI oversight is increasing. 79% of S&P 500 firms now disclose board committee oversight of AI, up from 72%.

What this means: Companies are taking AI governance seriously behind closed doors, even as they become more careful about what they say publicly. The gap between internal governance and external disclosure is widening - and that gap is where investment opportunities hide.

The Five-Step AI Washing Detection Framework

How do you separate companies with real AI capabilities from those that are AI washing? We developed a five-step framework that any investor can apply when reading a 10-K filing:

1. Check the Business Description

Does AI appear in the 10-K business section with specific products or services?

Green Flag:Named AI products with described functionality and target markets

Red Flag:Vague references to leveraging AI without naming what it does

2. Follow the Money

Is AI reflected in capital expenditure, R&D spending, or segment reporting?

Green Flag:Dedicated AI CapEx line items, AI-specific R&D increases, or new AI-related segments

Red Flag:Flat or declining R&D despite AI claims; no CapEx allocation

3. Read the Risk Factors

Are AI risks specific and operational, or generic and boilerplate?

Green Flag:Risks tied to specific AI products: bias in credit models, FDA approval for diagnostic AI

Red Flag:Generic risks like 'AI could disrupt our industry' without operational specifics

4. Compare Filings Year-over-Year

How has AI language evolved from last year to this year?

Green Flag:Progression from aspirational to operational language; new metrics and KPIs added

Red Flag:Same copy-pasted AI paragraphs with no new substance or specifics

5. Cross-Reference Earnings Calls

Do earnings call AI claims match the risk factors and financial data?

Green Flag:Consistent narrative between calls and filings; CFO can quantify AI impact

Red Flag:CEO talks AI constantly on calls, but filings show no investment or revenue attribution

Putting It All Together

A company that scores green flags across all five steps is likely a genuine AI adopter. A company with three or more red flags is probably AI washing. Most companies fall somewhere in between - and that is where careful 10-K reading gives you an edge over investors who only listen to earnings calls.

Pro tip: Compare a company's current 10-K AI language with their filing from two years ago. If the language evolved from generic to specific (named products, dollar figures, dedicated teams), that is a genuine adoption trajectory. If the same boilerplate appears year after year, it is theater.

The DEI Warning: What History Tells Us About Disclosure Fads

Want to know what happens when companies add buzzwords to SEC filings without substance? Look at what happened with DEI disclosures. In 2024, 90% of S&P 500 companies mentioned DEI in their 10-K filings. By 2025, that number collapsed to 34%.

DEI Disclosure Collapse

  • S&P 500 DEI mentions: 90% (2024) 34% (2025)
  • Use of "DEI" acronym dropped 68% in filings
  • Board diversity data disclosure: 90.5% 60.4%
  • DEI-linked executive comp: 55% 34%
  • 200+ companies scrubbed "diversity" and "equity" from 10-Ks

Lesson for AI Disclosures

  • Disclosure trends follow political and regulatory winds
  • Companies add buzzwords when rewarded, remove when punished
  • Substantive programs survive; performative ones disappear
  • AI disclosures could follow the same boom-bust pattern
  • Companies with real AI will keep disclosing; AI washers will stop

The AI Disclosure Prediction

If the SEC follows through on its enforcement priorities, expect AI disclosures to undergo their own "DEI moment" within 2-3 years. Companies with genuine AI operations will continue expanding their disclosures (with more specifics, metrics, and board oversight). Companies that were AI washing will quietly remove the language - just like they did with DEI. The companies that keep disclosing in 2028 will be the real AI winners.

The Bottom Line: Follow the Money, Not the Words

The AI disclosure explosion is the most dramatic shift in SEC filing language since the post-2008 financial risk disclosures. But volume does not equal substance. As the SEC tightens enforcement and investors get savvier, the gap between companies that use AI and companies that talk about AI will become the most important analytical distinction of this decade.

Three Takeaways

1

The Numbers Do Not Lie

84% of S&P 500 companies mention AI, but only 40% provide substantive disclosures. That 44% gap represents companies where AI is marketing, not operations.

2

The SEC Is Watching

AI washing is now an official enforcement priority. Companies making misleading AI claims face the same scrutiny that greenwashers faced in ESG. The enforcement actions are coming.

3

Your Edge Is in the Filings

Use our five-step framework to read 10-K filings critically. Compare business descriptions to CapEx, cross-reference earnings calls with risk factors, and track disclosure evolution year-over-year.

Want to Analyze SEC Filings Like a Pro?

We break down 8-K material events, 10-K risk factors, and earnings disclosures for the companies that matter most. No buzzwords, just analysis.

Disclaimer: This analysis is for educational and informational purposes only. It is not investment advice. The statistics cited are sourced from The Conference Board, Harvard Law School Forum on Corporate Governance, CAQ, and other public research organizations. Always do your own research and consult with a qualified financial advisor before making investment decisions.

All data sourced from public SEC filings and published research. Analysis and opinions are those of the author. Past performance does not guarantee future results.