How to Find Companies With Active AI Budgets Using Only Public Data

There is too much AI money in market right now to waste time on fake demand. Gartner says worldwide AI spending will reach $2.5 trillion in 2026, with AI infrastructure alone at $1.37 trillion. BCC Research says major technology companies are investing $650 billion annually in AI infrastructure. That money is not evenly distributed. Some accounts have real, approved budgets and operational mandates. Others are still running workshops, pilots, and executive theater.

If you can tell the difference from public data before the first call, your pipeline gets better fast.

AI exploration vs. AI budget commitment

Exploration shows up as labs, pilots, proofs of concept, hackathons, and vague language like “evaluating use cases” or “testing productivity tools.” These companies may buy small software subscriptions, but they have not locked in serious spend.

Committed AI budget looks different. It shows up in earnings language, investor decks, annual reports, hiring plans, infrastructure announcements, and named programs with executive owners. The company stops talking about possibilities and starts talking about deployment, operating leverage, capacity constraints, backlog, training targets, acquisitions, and dedicated teams.

This matters because sales motion changes. Explorers need education. Budgeted buyers need implementation, governance, integration, security, data, and scale.

The five public signals of committed AI budget

1) Formal budget reclassification or capital allocation language. The cleanest signal is when AI moves out of innovation talk and into capital planning. Alphabet did this in public. On June 1, 2026, it announced an $80 billion equity capital raise to expand AI infrastructure and compute. The related SEC filing says the raise is part of Alphabet’s plan to fund “world-class AI compute infrastructure” and notes 2026 capex of $180 billion to $190 billion. That is not AI curiosity. That is balance-sheet-level commitment.

2) Dedicated AI headcount. Companies do not staff 1,000-plus specialists for a side project. Accenture said in its Q2 FY2026 conference call transcript that it now has more than 85,000 AI and data professionals, already above its goal of 80,000 by the end of fiscal 2026. It also said 192,000 employees completed its agentic AI fundamentals program. That is what operational budget looks like: trained workforce, scale targets, and hiring tied to delivery.

3) Infrastructure investment press releases. When a company announces in-country compute, capacity expansion, or AI-linked infrastructure, budget is usually already approved. Microsoft’s April 2026 announcement committed $10 billion in Japan from 2026 through 2029 for AI infrastructure, cybersecurity, and workforce development. It included expansion of in-country infrastructure and training for more than one million workers by 2030. Again: not experimentation. This is geographic capital deployment.

4) Named AI initiatives with executive ownership. Real budgets produce named platforms, products, or programs. JPMorgan’s public materials show this clearly. In its 2025 line-of-business shareholder letter, the bank said more than 65,000 CIB colleagues actively use LLM Suite, its internal generative AI platform, and that over 90% of engineers use AI code assistants. Named platform plus user base plus scaled internal adoption is a strong budget signal. The bank’s AI research organization is also public at JPMorgan AI Research, with focus areas including foundation models, multimodal document processing, and AI trust and safety.

5) Operational deployment language replacing pilot language. Look for verbs like implemented, scaled, deployed, in production, and used today. CBRE’s 2025 annual report says the company is using AI today for product differentiation and operational efficiency. It also says critical infrastructure accounted for about 14% of core EBITDA in 2025, up from about 3% in 2021, and explicitly ties much of that work to the “massive investment in Artificial Intelligence.” That is a useful cross-signal: AI language tied to current operations and measurable business mix.

Where to find these signals

You do not need paid intent data to find most of this. You need disciplined source coverage.

1) Investor relations newsrooms and earnings pages. Start with the company’s IR site and search for AI, capex, infrastructure, backlog, and hiring. Example: Alphabet Investor Relations, Accenture Investor Relations, JPMorgan investor relations.

Use search strings like site:company.com/investor AI infrastructure capex and site:company.com/investor earnings transcript AI deployment.

2) SEC filings. 10-Ks, 10-Qs, 8-Ks, and free writing prospectuses often carry the hard language. Use SEC EDGAR. Search for company name artificial intelligence filetype:pdf site:sec.gov or site:sec.gov company name AI infrastructure 8-K.

3) Annual reports and investor day decks. These documents surface strategy, executive ownership, and funding priorities better than headlines. GE HealthCare’s 2024 annual report filed in 2025 says the company increased AI-enabled FDA authorizations from 58 to 85 in one year, is developing CareIntellect cloud solutions, and is working on foundation models. That combination of regulatory throughput, product naming, and platform development is what you want.

4) Press releases. PR is useful when it includes specific dollars, headcount, geographies, or deployment milestones. GE HealthCare’s 2025 AI Innovation Lab announcement says the company is moving toward a “cloud-first, AI-powered, and software-enabled company” and had reached 100 AI-enabled device authorizations.

5) Careers pages. Hiring tells you whether the company is operationalizing. Search for titles like LLM Operations Engineer, AI Product Management Practitioner, Enterprise AI R&D, AI Trust, and AI Governance. Examples: Accenture has a public posting for LLM Operations Engineer and another for AI Product Management Practitioner. GE HealthCare has posted roles including Senior Product Director, Enterprise AI R&D and Senior AI Application Engineer, Enterprise AI.

Useful search strings: site:company.com/careers "LLM Operations Engineer", site:company.com/careers "AI governance", site:company.com/careers "Enterprise AI".

A real worked example: GE HealthCare

If you want one non-hyperscaler account showing active AI budget signals right now, GE HealthCare is a strong example.

First, its annual report says it is developing cloud-based solutions like CareIntellect and foundation models, and that its AI-enabled FDA authorizations rose from 58 to 85 in one year. Second, its 2025 AI Innovation Lab release says it reached 100 AI-enabled device authorizations and is pushing a cloud-first, AI-powered strategy. Third, its careers footprint includes dedicated roles such as Senior Product Director, Enterprise AI R&D and Technology Director, Connected Health and Artificial Intelligence.

That stack of signals matters because it spans regulatory output, named products, platform strategy, and specialized hiring. That is what a real buying account looks like before procurement ever talks to you.

What to do with this

Build a simple AI budget score for target accounts. Give one point each for: investor-language capex signal, dedicated AI headcount, infrastructure announcement, named AI initiative, and operational deployment language. If an account scores 4 or 5, work it. If it scores 0 or 1, deprioritize it until the signals change.

This is how you stop chasing AI hype and start chasing funded demand.

If you want these signals surfaced for you every week instead of hunting them manually, SalesInt’s paid tier includes Signal Watch, our Monday brief tracking budget-grade buying signals across industries before the rest of the market catches up.