Login Get a Demo
01 Research · 2026

AI Supply Chain Statistics 2026

Twenty figures on shadow AI, vendor AI, open models, agents and regulation. Each one is dated, linked to its primary source and labeled when a vendor paid for the research.

The short answer

What do AI supply chain statistics show in 2026?

Last updated
Read time
8 min read
From
Cranium AI
Share on LinkedIn

AI supply chain statistics measure how much of an organization’s AI is built, bought or downloaded from outside, and how well it is tracked.

In 2026, public sources show a third party involved in 48% of breaches, third-party systems making up 39% of AI at UK financial firms, and 45% of employees regularly using AI on corporate devices.

Key takeaways
  • A third party was involved in 48% of breaches in Verizon’s 2026 dataset, up 60% in a year.
  • At UK financial firms, third-party systems are 39% of all AI and 53% of high-materiality AI.
  • 45% of employees regularly use AI on corporate devices, and 67% of AI users do it through non-corporate accounts.
  • Every figure links to its primary source. Vendor-funded research is labeled, and none of it is Cranium AI customer data.
02 At a glance

The AI supply chain, by the numbers

Four figures frame the rest. Each is reported as its source published it, with the scope stated. The full list of twenty follows, grouped under five findings.

01

48% of breaches involved a third party

Up 60% from the previous year’s dataset. Covers all breaches, not only AI‑related ones.

Verizon, 2026 DBIR
03

45% of employees use AI at work

Regular users of AI on corporate devices, authorized or not, up from 15% a year earlier.

Verizon, 2026 DBIR
03 Shadow AI

Shadow AI statistics: how much AI runs unrecorded?

An inventory only helps if it matches what people use. These figures measure the gap between the two. For the definition and the four ways it gets in, see What Is Shadow AI?

FigureWhat it measuresSource
Stat 0145%Employees who are regular users of AI on their corporate devices, authorized or not, up from 15% a year earlier. Measured in the report’s 2026 dataset.Verizon Business, 2026 Data Breach Investigations Report, Executive SummaryPublished May 19, 2026 · Source 1
Stat 0267%AI users who reach AI services through non-corporate accounts on their corporate devices, a slight decrease from the year before.Verizon Business, 2026 Data Breach Investigations Report, Executive SummaryPublished May 19, 2026 · Source 1
Stat 031 in 5Organizations that reported a breach due to shadow AI. High levels of shadow AI added an average of $670,000 to breach costs. Based on 600 organizations breached between March 2024 and February 2025.IBM, Cost of a Data Breach Report 2025, press releasePublished July 30, 2025 · Source 2Vendor-sponsored study
Stat 0463%Breached organizations in the same study that had no AI governance policy or were still developing one. Of those with a policy, 34% ran regular audits for unsanctioned AI.IBM, Cost of a Data Breach Report 2025, press releasePublished July 30, 2025 · Source 2Vendor-sponsored study
04 Vendor AI

Third-party AI risk statistics: how much AI is bought?

Most organizations buy more AI than they build, and understand it through what the vendor tells them. The third-party AI risk management hub covers what to do about it.

FigureWhat it measuresSource
Stat 0548%Breaches that involved a third party, up 60% from the previous year’s dataset. This covers all breaches in the report, not only AI‑related ones.Verizon Business, 2026 Data Breach Investigations Report, Executive SummaryPublished May 19, 2026 · Source 1
Stat 0639%Share of all AI systems reported by UK financial services firms that are third-party implementations, up from 33% in 2024. Based on 196 survey responses.Bank of England and Financial Conduct Authority, Artificial intelligence in UK financial services 2026Published October 26, 2026 · Source 3
Stat 0753%Share of high-materiality AI systems at those firms that are third-party implementations, up from 35% in 2024.Bank of England and Financial Conduct Authority, Artificial intelligence in UK financial services 2026Published October 26, 2026 · Source 3
Stat 0881%Respondents that rely on vendor assurance and contractual arrangements to understand their high-materiality third-party AI systems. 24% rely on deep knowledge of those systems, against 77% for systems built in-house.Bank of England and Financial Conduct Authority, Artificial intelligence in UK financial services 2026Published October 26, 2026 · Source 3
Stat 0972%Financial institutions that are only partially aware of which of their vendors use AI. A further 9% have not assessed it at all. The sample size is not published.Ncontracts, 2026 State of Third-Party Risk Management SurveyPublished June 4, 2026 · Source 4Vendor survey
Stat 1057.9%EU enterprises using AI that bought it as ready-to-use commercial software, the most common route. 21.8% developed their own. 2024 data, enterprises with 10 or more employees.Eurostat, The use of artificial intelligence (AI) technologies in the European Union: Key results, 2026 editionPublished 2026 · Source 5
05 Models and data

Open models and data: how fragile are the inputs?

Models, training data and the code that AI writes are supply chain components too. An AI‑BOM is how you record which ones a system depends on.

FigureWhat it measuresSource
Stat 11250Malicious documents that were enough to backdoor language models ranging from 600 million to 13 billion parameters. The backdoor tested was narrow, and the authors say it is unclear whether the pattern holds for larger models.Anthropic, UK AI Security Institute and The Alan Turing Institute, A small number of samples can poison LLMs of any sizePublished October 9, 2025 · Source 6
Stat 1221.7%Average share of software packages recommended by open-source code models that do not exist. For commercial models the figure is at least 5.2%. Measured across 16 models and 576,000 code samples.Spracklen et al., We Have a Package for You! A Comprehensive Analysis of Package Hallucinations by Code Generating LLMs, USENIX Security 2025Published March 2, 2025 (v3) · Source 7
Stat 13205,474Unique names of non-existent packages the models produced in the same study. The paper examines how such names open the door to package confusion attacks.Spracklen et al., We Have a Package for You! A Comprehensive Analysis of Package Hallucinations by Code Generating LLMs, USENIX Security 2025Published March 2, 2025 (v3) · Source 7
06 Agents

Agents and tools: how far has adoption run?

An agent is a model plus tools and permissions, and each connection is a component someone else built. Our guide to agentic AI security covers the risks OWASP names.

FigureWhat it measuresSource
Stat 1480%UK financial services respondents with AI agents deployed; a further 14% plan to deploy them within 12 months. Only 2% report agents with broad autonomy.Bank of England and Financial Conduct Authority, Artificial intelligence in UK financial services 2026Published October 26, 2026 · Source 3
Stat 1580%+Fortune 500 companies using active AI agents built with Microsoft’s low-code and no-code tools, measured over the last 28 days of November 2025.Microsoft Security Blog, 80% of Fortune 500 use active AI AgentsPublished February 10, 2026 · Source 8Vendor telemetry
Stat 1627%Share attributed to compromised APIs, applications or plug-ins among the causes of breaches targeting AI models or applications. Cloud misconfigurations affecting AI workloads were level at 27%. More than 20% of organizations reported such a breach.IBM, Cost of a Data Breach Report 2026, press releasePublished July 29, 2026 · Source 9Vendor-sponsored study
Stat 1792%Organizations reporting an AI-related breach that lacked proper AI access controls. Roughly one in five organizations in the study reported an AI-related breach.IBM X-Force, AI-powered adversaries and the enterprise risk challengePublished July 29, 2026 · Source 10Vendor-sponsored study
07 Regulation

Regulation and evidence: who is asking for proof?

Boards disclose AI risk, and regulators have started to name the buyer’s duties. See AI compliance frameworks, compared for the dates and the three frameworks side by side.

FigureWhat it measuresSource
Stat 1883%S&P 500 companies that disclosed AI as a risk in 2025, up from 12% in 2023. Disclosure data as of December 2025.The Conference Board, In the S&P 500, Disclosure of AI Risks Surges from 12% to 83%Published April 22, 2026 · Source 11
Stat 19EUR 15MMaximum EU AI Act fine for breaching the obligations of deployers of high-risk AI: EUR 15 million or 3% of worldwide annual turnover, whichever is higher. Rules for Annex III high-risk systems apply from December 2, 2027.Regulation (EU) 2024/1689 (AI Act), Articles 26, 99(4) and 113, as amended by Regulation (EU) 2026/1744Published July 27, 2026 (consolidated text) · Source 12
Stat 2019Critical ICT third-party providers on the list designated by the European Supervisory Authorities under the Digital Operational Resilience Act, which has applied to EU financial entities since January 17, 2025.European Supervisory Authorities (EBA, EIOPA, ESMA), List of designated critical ICT third-party providers under DORAPublished November 18, 2025 · Source 13
08 Methodology

How these statistics were chosen

This page is a synthesis of public sources. Cranium AI ran no new survey and used no customer data. Every figure was re-checked against its primary source on November 17, 2026.

01

Primary sources only

Each figure links to the organization that produced it: a regulator, a statistics office, a research paper or the publisher’s own release. Nothing is taken from a news summary.

02

Reported as published

We keep each source’s wording and scope. Where a figure covers one sector or one country, the row says so instead of implying a global number.

03

Vendor research is labeled

Surveys, telemetry and studies paid for by a vendor carry a visible label. They can still be useful, and you should know who funded them.

04

What we left out

Analyst forecasts, figures we could not confirm at the source, and research from companies we are compared with on this site.

How to cite this page. Cranium AI, “AI Supply Chain Statistics 2026,” last updated November 17, 2026, https://cranium.ai/ai-supply-chain-statistics/. When you quote a single figure, credit the original source listed beside it as well: this page is a compilation, and the numbers belong to the organizations that published them.
09 Sources

Numbered sources

Thirteen sources, each opened on November 17, 2026. The right-hand column shows which statistics each one supports.

SourcePublishedStatistics
Source 1
Verizon Business, 2026 Data Breach Investigations Report, Executive Summary
May 19, 2026Stat 01, 02, 05
Source 2
IBM, Cost of a Data Breach Report 2025, press release
Vendor-sponsored study
July 30, 2025Stat 03, 04
Source 3
Bank of England and Financial Conduct Authority, Artificial intelligence in UK financial services 2026
October 26, 2026Stat 06, 07, 08, 14
Source 4
Ncontracts, 2026 State of Third-Party Risk Management Survey
Vendor survey
June 4, 2026Stat 09
Source 5
Eurostat, The use of artificial intelligence (AI) technologies in the European Union: Key results, 2026 edition
2026Stat 10
Source 6
Anthropic, UK AI Security Institute and The Alan Turing Institute, A small number of samples can poison LLMs of any size
October 9, 2025Stat 11
Source 7
Spracklen et al., We Have a Package for You! A Comprehensive Analysis of Package Hallucinations by Code Generating LLMs, USENIX Security 2025
March 2, 2025 (v3)Stat 12, 13
Source 8
Microsoft Security Blog, 80% of Fortune 500 use active AI Agents
Vendor telemetry
February 10, 2026Stat 15
Source 9
IBM, Cost of a Data Breach Report 2026, press release
Vendor-sponsored study
July 29, 2026Stat 16
Source 10
IBM X-Force, AI-powered adversaries and the enterprise risk challenge
Vendor-sponsored study
July 29, 2026Stat 17
Source 11
The Conference Board, In the S&P 500, Disclosure of AI Risks Surges from 12% to 83%
April 22, 2026Stat 18
Source 12
Regulation (EU) 2024/1689 (AI Act), Articles 26, 99(4) and 113, as amended by Regulation (EU) 2026/1744
July 27, 2026 (consolidated text)Stat 19
Source 13
European Supervisory Authorities (EBA, EIOPA, ESMA), List of designated critical ICT third-party providers under DORA
November 18, 2025Stat 20
10 The full report

Read the 2026 report

These statistics accompany a 15-page research report that adds the analysis: what the findings mean, a field playbook and a self-assessment.

The AI Supply Chain Risk Report 2026 cover Research Report
Free download · PDF · 15 pages

The AI Supply Chain Risk Report 2026

Built for the security, risk and governance leaders who own AI supply chain risk.

  • Five findings from 26 published sources
  • A five-move field playbook
  • A 10‑question self‑assessment for your AI supply chain
Download the report
11 Where Cranium AI fits

Where Cranium AI fits

These figures describe a market, not our customers. Cranium AI works on the part a buyer can act on, through the five moves of the AI Trust Loop. It does not see a model a vendor runs in its own infrastructure until the vendor discloses it, and it does not replace a TPRM suite.

  1. Discover

    Build the inventory the shadow AI figures say is missing: an AI‑BOM from your code repositories, and a Shadow AI view from connected SIEM activity.

  2. Observe

    Guardian inspects traffic between your AI applications and the models they call, and shows the tool calls your agents make.

  3. Govern

    Map each AI system to the EU AI Act, NIST AI RMF and ISO 42001, with attestations backed by citations.

  4. Secure

    Arena tests the AI applications you build against adversarial attacks before they reach production.

  5. Prove

    Request an AI Card from each vendor and keep the returned cards in My Vendors. AI‑BOMs export as JSON, CycloneDX, SPDX and XLSX.

12 Keep going

Related reading.

13 Questions

Frequently asked questions.

Straight answers, kept current. Last updated .

Ask us anything →
What percentage of data breaches involve a third party?

Verizon’s 2026 Data Breach Investigations Report found that breaches with third-party involvement rose 60% from the previous year’s dataset and reached 48% of all breaches. The figure covers every breach in the dataset, not only those involving AI. The report analyzed more than 22,000 confirmed breaches.

How much enterprise AI comes from third parties?

In the Bank of England and Financial Conduct Authority’s 2026 survey of UK financial services firms, third-party implementations made up 39% of all reported AI systems and 53% of high-materiality systems. Across the EU, Eurostat reports that 57.9% of enterprises using AI in 2024 bought ready-to-use commercial software, the most common way of acquiring it.

What are the latest shadow AI statistics?

Verizon’s 2026 report found that 45% of employees are regular users of AI on corporate devices, up from 15% a year earlier, and that 67% of AI users reach those services through non-corporate accounts. In IBM’s 2025 study, one in five organizations reported a breach due to shadow AI, and 63% of breached organizations had no finished AI governance policy.

How many poisoned documents does it take to backdoor a language model?

Research published in October 2025 by Anthropic, the UK AI Security Institute and The Alan Turing Institute found that 250 malicious documents were enough to backdoor language models ranging from 600 million to 13 billion parameters. The backdoor tested was a narrow one, and the authors say it is unclear whether the pattern holds for larger models.

How many companies disclose AI as a risk?

The Conference Board reported in April 2026 that the share of S&P 500 companies disclosing AI as a risk rose from 12% in 2023 to 83% in 2025, based on disclosure data as of December 2025. Disclosure is not the same as control: it tells investors the risk exists, not that it is managed.

Where do these AI supply chain statistics come from?

From thirteen public sources: regulators, a statistics office, peer-reviewed research and publishers’ own releases, each opened on November 17, 2026. Cranium AI ran no survey and used no customer data. Research paid for by a vendor is labeled, and analyst forecasts and figures we could not confirm at the source are left out.

Can I cite these statistics?

Yes. Cite the original source for any single figure, since the number belongs to the organization that published it, and cite this page when you refer to the compilation. A suggested citation appears in the methodology section, and every row links to the source so you can check the figure before you use it.

14 See it for yourself

The numbers describe the market.
See your own supply chain.

See how Cranium AI discovers, observes, governs, secures and proves every model, agent and vendor in your enterprise.