What Companies Say vs. What They Do
The most revealing thing about enterprise AI is not any single data point — it is the gap between multiple sources measuring the same behavior. When we cross-reference what companies report in SEC filings with who they actually hire, how much they spend, and what surveys say, the narrative significantly outpaces execution.
SEC 10-K filings are legally required disclosures. They reflect what companies need investors to believe. McKinsey surveys reflect aspirational strategy. BLS hiring data reflects operational reality. LinkedIn job postings reflect what skills companies are actually willing to pay for. These four datasets measure the same phenomenon from four different angles — and they produce very different numbers.
Industry-by-Industry Gap Analysis
| Industry | 10-K AI Mention Rate SEC EDGAR 2024 |
Stated Adoption McKinsey Q1 2026 |
Hiring Signal BLS + LinkedIn 2025 |
Execution Gap |
|---|---|---|---|---|
| Manufacturing | 71% | 29% | 12% | 59pt gap |
| Healthcare | 68% | 62% | 15% | 53pt gap |
| Retail & Consumer | 65% | 45% | 20% | 45pt gap |
| Energy & Utilities | 61% | 38% | 17% | 44pt gap |
| Financial Services | 78% | 79% | 48% | 30pt gap |
| Technology | 91% | 88% | 67% | 24pt gap |
Execution gap = 10-K mention rate minus hiring signal (BLS AI-specialist hiring growth). Lower gap = closer alignment between stated strategy and operational execution. Sources: SEC EDGAR full-text search Q4 2024; McKinsey Global Survey on AI Q1 2026; BLS JOLTS Nov 2025; LinkedIn Workforce Report 2025.
Manufacturing has the highest narrative-to-execution gap of any major sector. This is structural: 71% of manufacturers mention AI in SEC filings — largely driven by investor expectation and competitive pressure — but only 12% have materially increased AI-specialist hiring. The likely explanation is that manufacturers are using AI as a compliance narrative while actual AI implementation remains limited by legacy infrastructure and workforce transition costs that don't appear in filings.
The SEC Filing Signal
The number of Fortune 500 companies mentioning "artificial intelligence" or "machine learning" in SEC 10-K filings grew from 57% in 2022 to 73% in 2024 (SEC EDGAR full-text search). This is a 16-point increase in two years — the fastest growth rate for any technology disclosure category in SEC history.
However, the quality of the mention matters. AIStackHub analysis of a 50-company sample found that 61% of 10-K AI mentions were in forward-looking statements or risk-factor language ("if we fail to adapt to AI..."), not in descriptions of current capabilities. Only 39% described specific AI deployments in production. This distinction is invisible in aggregate statistics but critical for understanding actual adoption depth.
The Hiring Signal
BLS data shows AI-related job titles grew 28% year-over-year in 2025. LinkedIn's Workforce Report shows "AI/ML Engineer" job postings grew 32% in the same period. These are strong numbers — but they are concentrated. Stanford HAI's 2026 Index found that 72% of AI-related job growth occurred at just 50 companies, most of them technology companies that were already AI-native.
For the other 450 Fortune 500 companies, the hiring signal is weak relative to the filing narrative. A company can mention AI prominently in investor communications while adding only a handful of AI specialists — or none.
The McKinsey Survey Signal
McKinsey's Q1 2026 Global Survey on AI reports 65% of enterprises have deployed generative AI in at least one business function, up from 33% in early 2023. This is a genuine and rapid adoption signal. But it coexists with the hiring gap because many AI deployments are via SaaS tools that require no internal AI expertise — deploying ChatGPT Enterprise or Copilot does not require hiring an ML engineer.
This explains the apparent paradox: adoption is real and growing, but the organizational depth of that adoption is shallow. A company using AI features in its CRM or email platform is "AI-deployed" by survey standards, but shows no AI-specialist hiring signal in BLS data.
Methodology
SEC EDGAR source: Full-text search of Fortune 500 2024 annual 10-K filings for terms "artificial intelligence," "machine learning," and "generative AI." Measured as binary (mention or no mention). Sample confirmation of 50 filings for mention quality analysis (forward-looking vs. operational). Access date: April 2026.
BLS source: Bureau of Labor Statistics Job Openings and Labor Turnover Survey (JOLTS) plus Occupational Employment and Wage Statistics (OEWS). AI-related job titles defined as: AI Engineer, ML Engineer, Data Scientist (AI focus), AI Product Manager, LLM Engineer. Growth measured YoY 2024→2025. Access date: May 2026.
McKinsey source: McKinsey Global Survey on AI, Q1 2026 (publicly available summary). Industry breakdowns from published report tables. "Deployed" defined as McKinsey's "using AI in at least one function." Access date: April 2026.
LinkedIn source: LinkedIn Workforce Report 2025 (publicly available). AI/ML job posting growth rates and geographic concentration data. Access date: March 2026.
Gap calculation: Execution gap = 10-K mention rate minus BLS AI-specialist hiring growth rate. This is a simplified proxy; the underlying drivers include SaaS AI tool adoption (which inflates stated adoption without generating hiring) and industry-specific structural barriers.