AI Governance Statistics 2026: The Global Data Lawmakers, Executives, and Researchers Are Watching

AI Governance Statistics 2026

By Axis Intelligence Research & Sarah Mitchell | Last updated: June 7, 2026 | Next scheduled update: Q3 2026 (September) | Download CSV ↓

Quick Answer: As of June 2026, 47 countries have enacted active AI legislation, yet only 12 have functioning enforcement mechanisms — a gap that produced 156 documented AI regulatory enforcement actions in 2025 alone, up 263% from 43 in 2024 (Stanford HAI, 2026 AI Index Report). Global spending on AI governance platforms will reach $492 million in 2026 and surpass $1 billion by 2030, according to Gartner.

The distance between writing an AI policy and enforcing one has never been wider. In 2016, virtually no country had AI-specific legislation. By 2026, the OECD AI Policy Observatory tracks over 1,000 active policy initiatives across more than 70 jurisdictions — and the world’s most consequential AI law, the EU AI Act, begins full enforcement for high-risk systems on August 2, 2026.

Yet the regulatory machinery is outpaced by deployment. McKinsey’s 2026 AI Trust Maturity Survey found that the average organizational responsible AI (RAI) maturity score stands at just 2.3 out of 5 — and only about 30% of organizations have reached maturity level 3 or higher in strategy, governance, and agentic AI controls. The gap between AI capability and AI governance has become the defining risk of the decade.

This report compiles verified statistics from primary sources — Stanford HAI, OECD, Gartner, McKinsey, NIST, and the EU AI Office — to give policymakers, compliance teams, journalists, and researchers the most complete picture available of where global AI governance stands as of mid-2026.

---

Key Findings

• Enforcement is accelerating sharply. Documented AI regulatory enforcement actions jumped from 43 in 2024 to 156 in 2025, a 263% single-year increase, according to the 2026 Stanford HAI AI Index Report.
• The legislation-enforcement gap is structural, not temporary. Of the 47 countries with active AI legislation, only 12 have established enforcement mechanisms — leaving 74% of AI laws with no practical teeth as of mid-2026.
• Corporate governance roles are formalizing. AI-specific governance roles grew 17% in 2025, and the share of businesses with no responsible AI policies fell from 24% to 11%, per Stanford HAI’s 2026 AI Index.
• Model transparency is going backward. The Foundation Model Transparency Index average dropped from 58/100 in 2024 to 40/100 in 2025 — its steepest decline since the index launched. Of 95 notable models released in 2025, 80 shipped with no published training code.
• The governance market is now a billion-dollar race. Gartner projects AI regulation will quadruple and cover 75% of global economies by 2030, driving $1 billion in total compliance spend — with $492 million already expected in 2026.
• The US public trust deficit is measurable. Just 31% of U.S. citizens trust their government to regulate AI effectively — the lowest score among surveyed nations (except China at 27%). EU citizens stand at 53%, per Stanford HAI 2026.

---

Global Legislative Landscape

How many countries have AI laws?

The legislative surge is real but uneven. According to the 2026 Stanford HAI AI Index Report, 47 countries had active AI legislation as of early 2026 — but only 12 had established enforcement mechanisms. That 74% enforcement gap represents the core structural challenge of AI governance globally.

Region	Data Localization Measures (through 2024)	Status
East Asia & Pacific	77	Highest volume globally
Sub-Saharan Africa	71	Fast-moving; framework-driven
Europe & Central Asia	66	EU AI Act leads; extraterritorial reach
North America	3	Sectoral approach; federal framework absent

Source: Stanford HAI, 2026 AI Index Report, Policy and Governance chapter. https://hai.stanford.edu/ai-index/2026-ai-index-report/policy-and-governance

The North America figure — just 3 data localization measures compared to 77 in East Asia — illustrates the philosophical divide in regulatory approach. The US model has prioritized sectoral regulation and executive action over binding horizontal frameworks, while the EU has pursued comprehensive legislation that applies regardless of where a company is headquartered.

AI policy initiatives by jurisdiction

The OECD AI Policy Observatory, the most comprehensive public database of national AI policies, tracks over 1,000 active policy initiatives across more than 70 countries and territories. The database — updated continuously — allows filtering by policy instrument, target group, and country.

Metric	Figure	Source
Countries with active AI legislation	47	Stanford HAI, 2026 AI Index Report
Countries with enforcement mechanisms	12	Stanford HAI, 2026 AI Index Report
AI policy initiatives tracked globally	1,000+	OECD AI Policy Observatory
Jurisdictions in OECD AI database	70+	OECD.AI Dashboard
AI enforcement actions in 2025	156 (vs. 43 in 2024)	Stanford HAI, 2026 AI Index Report
Compliance cost variation between jurisdictions	Up to 8×	Stanford HAI, 2026 AI Index Report

Sources: Stanford HAI AI Index 2026 (https://hai.stanford.edu/ai-index/2026-ai-index-report); OECD.AI Policy Navigator (https://oecd.ai/en/dashboards/national)

AI sovereignty infrastructure

Beyond laws, countries are competing on AI infrastructure as a governance and sovereignty strategy. According to Stanford HAI’s 2026 Policy and Governance chapter, Europe and Central Asia expanded state-backed AI supercomputing clusters from 3 to 44 between 2018 and 2025. South Asia, Latin America, and the Middle East and North Africa have each only reached 2–8 clusters.

EU AI Act — The World’s First Binding AI Law

Enforcement timeline and penalties

The EU AI Act entered into force on August 1, 2024 and is phasing in obligations across four key dates. Its penalty structure is the most aggressive in any technology regulation to date.

Date	What Becomes Enforceable
February 2, 2025	Prohibited AI practices bans; AI literacy requirements
August 2, 2025	GPAI model obligations; governance infrastructure
August 2, 2026	High-risk AI system obligations (most demanding deadline)
August 2, 2027	AI systems embedded in regulated products

Source: EU AI Act Implementation Timeline, artificialintelligenceact.eu (https://artificialintelligenceact.eu/implementation-timeline/)

The August 2, 2026 deadline is the most operationally consequential date in AI compliance history. High-risk AI systems — those used in biometric identification, critical infrastructure, education, employment, essential services, law enforcement, migration, and judicial processes — must complete conformity assessments, implement quality management systems, and register in the EU database before market placement.

Penalty structure

Violation Type	Maximum Penalty
Prohibited AI practices violations	€35 million or 7% of global turnover
High-risk system non-compliance	€15 million or 3% of global turnover
Incorrect information to authorities	€7.5 million or 1% of global turnover
GDPR overlap (personal data violations)	€20 million or 4% of annual turnover

Source: EU AI Act; legiscope.com analysis of penalty provisions (https://www.legiscope.com/blog/eu-ai-act-timeline-deadlines.html)

Corporate AI Governance Adoption

Governance roles and policies

The organizational picture is improving — but slowly. Stanford HAI’s 2026 AI Index Report, drawing on enterprise survey data, found:

Metric	2024	2025	Change
Businesses with no responsible AI policies	24%	11%	−13 pp
AI-specific governance roles growth	—	+17%	YoY
Organizations with no regulatory influence cited	17%	12%	−5 pp
GDPR cited as regulatory influence	65%	60%	−5 pp
ISO/IEC 42001 cited as regulatory influence	New entry	36%	—
NIST AI RMF cited as regulatory influence	New entry	33%	—

Source: Stanford HAI, 2026 AI Index Report, Responsible AI chapter. https://hai.stanford.edu/ai-index/2026-ai-index-report/responsible-ai

The shift from GDPR to AI-specific frameworks is significant: ISO/IEC 42001 and NIST AI RMF were new entries in 2025, yet immediately appeared in the top-five regulatory influences. This reflects the practical reality that GDPR alone is no longer sufficient to cover the risk surface of agentic AI.

AI governance market size

Metric	Figure	Source
AI governance platform spending, 2026	$492 million	Gartner, February 2026
AI governance platform spending, 2030	$1 billion+	Gartner, February 2026
Global economies covered by AI regulation, 2030	75%	Gartner, February 2026
Regulation quadrupling projection by 2030	4× current volume	Gartner, February 2026
Potential compliance cost reduction (effective governance)	20%	Gartner, February 2026

Source: Gartner, “Global AI Regulations Fuel Billion-Dollar Market for AI Governance Platforms,” February 17, 2026.

McKinsey AI Trust Maturity findings (2026)

McKinsey’s 2026 AI Trust Maturity Survey — which gathered responses from approximately 500 organizations between December 2025 and January 2026 — measured RAI maturity across five dimensions: strategy, risk management, data and technology, governance, and (new in 2026) agentic AI governance and controls.

Metric	2025	2026	Change
Average RAI maturity score (out of 5)	2.0	2.3	+0.3
Organizations at maturity level 3+ in strategy	—	~30%	—
Organizations at maturity level 3+ in governance	—	~30%	—
Organizations at maturity level 3+ in agentic AI controls	—	~30%	—

Source: McKinsey & Company, “State of AI Trust in 2026: Shifting to the Agentic Era,” March 2026. https://mckinsey.com/capabilities/tech-and-ai/our-insights/tech-forward/state-of-ai-trust-in-2026-shifting-to-the-agentic-era

The finding that only about one-third of organizations meet adequate governance maturity for agentic AI — at the exact moment organizations are scaling autonomous agent deployments — is the central governance risk of 2026. McKinsey’s earlier State of AI 2025 survey found 88% of organizations use AI in at least one business function, up from 78% the prior year.

AI Transparency and Incidents

Foundation Model Transparency Index

The Foundation Model Transparency Index (FMTI), published by the Stanford Center for Research on Foundation Models, is the most rigorous independent measure of corporate AI disclosure. Its 2025 edition produced the most alarming finding in the index’s history.

Year	Average FMTI Score (out of 100)
2023	37
2024	58
2025	40

Source: Stanford Center for Research on Foundation Models, “2025 Foundation Model Transparency Index,” December 2025; incorporated in Stanford HAI 2026 AI Index Report. https://hai.stanford.edu/ai-index/2026-ai-index-report/responsible-ai

The 18-point drop in a single year — after an improvement year in 2024 — signals that the most capable AI labs are moving in the wrong direction on transparency precisely as their commercial and strategic value rises. Notable individual scores: IBM at 95 (highest), xAI and Midjourney at 14 (lowest). Of the 95 most notable AI models released in 2025, 80 shipped with no published training code.

AI incidents

The AI Incident Database, tracked by Stanford HAI, recorded 362 documented AI incidents in 2025 — up from 233 in 2024, a 55% year-on-year increase.

Year	Documented AI Incidents	Change
2024	233	—
2025	362	+55%

Source: Stanford HAI, 2026 AI Index Report, Responsible AI chapter. https://hai.stanford.edu/ai-index/2026-ai-index-report/responsible-ai

US Congressional and Public Investment in AI Governance

Congressional AI attention

One of the most striking metrics in Stanford HAI’s 2026 Policy and Governance chapter tracks AI-related witnesses in US congressional hearings — a proxy for legislative urgency.

Year	AI Witnesses in Congressional Hearings	Industry Share	Academia Share
2017	5	13%	—
2025	102	37%	15%

Source: Stanford HAI, 2026 AI Index Report, Policy and Governance chapter. https://hai.stanford.edu/ai-index/2026-ai-index-report/policy-and-governance

Industry’s share of congressional AI witnesses nearly tripled from 13% to 37% between 2017 and 2025, making it the largest witness group. Academia’s share fell to 15%. This shift in who speaks to lawmakers reflects the broader commercialization of AI governance discourse — and raises questions about whose interests shape regulation.

Public investment vs. private

Region	AI Public Investment, 2013–2024	Top Country	Private Investment, 2025 (US only)
United States	~$20.4 billion	—	$285.9 billion
Europe	~$3.7 billion	UK ($1.6B)	—
Germany	$505 million	—	—
France	$320 million	—	—

Source: Stanford HAI, 2026 AI Index Report, Policy and Governance chapter. https://hai.stanford.edu/ai-index/2026-ai-index-report/policy-and-governance

The US invested $20.4 billion in AI-related contracts and grants over 11 years — against $285.9 billion in private AI investment in the single year 2025. The 14:1 ratio of private to public spending in a single year versus a decade of government investment is the clearest illustration of why AI governance capacity lags deployment.

NIST AI RMF — The US Governance Standard

The NIST AI Risk Management Framework (AI RMF), a voluntary but increasingly authoritative standard, has become the primary governance reference for US organizations and is mapped to by EU AI Act obligations globally. On April 7, 2026, NIST released a concept note for an AI RMF Profile on Trustworthy AI in Critical Infrastructure, extending the framework’s reach to power grids, financial systems, and healthcare infrastructure.

Key reference points for organizations:

• The AI RMF Generative AI Profile (NIST AI 600-1) maps 12 risk categories and over 200 specific actions to GenAI systems.
• The Cybersecurity Framework Profile for AI (NIST IR 8596) — initial draft released December 2025 — bridges NIST CSF 2.0 with the AI RMF.
• ISO/IEC 42001 is now cited by 36% of surveyed organizations as a compliance influence, up from zero in 2024 (Stanford HAI 2026).
• NIST’s AI Agent Standards Initiative, announced February 2026, will produce guidance on agentic AI identity, authorization, security, and monitoring — with an Agentic Interoperability Profile planned for Q4 2026.

Source: NIST, AI Risk Management Framework. https://www.nist.gov/itl/ai-risk-management-framework

Public Trust in AI Governance

Stanford HAI’s 2026 AI Index Report includes global public opinion data on trust in government AI regulation.

Country/Region	Trust Government to Regulate AI (%)
European Union	53%
Global average	—
United States	31%
China	27%

Source: Stanford HAI, 2026 AI Index Report, Public Opinion chapter. https://hai.stanford.edu/ai-index/2026-ai-index-report/public-opinion

That only 31% of US citizens trust their own government to regulate AI — lower than the EU’s 53% — reflects both the absence of a federal AI framework and a broader erosion of institutional trust in technology governance. The US score is also the second-lowest of surveyed nations, above only China at 27%.

The Axis Intelligence AI Governance Gap Index (AGGI™) — Q2 2026

To give researchers and compliance teams a single composite measure unavailable from any individual source, Axis Intelligence Research has constructed the AI Governance Gap Index (AGGI™), a cross-source metric that synthesizes four verified primary-source dimensions:

1. Legislative coverage score: Countries with enforcement mechanisms as a share of countries with active legislation (12/47 = 25.5%)
2. Corporate maturity deficit: Inverse of the share of organizations at RAI maturity level 3+ (100% − 30% = 70% governance-inadequate)
3. Transparency erosion rate: Year-on-year FMTI decline as a share of peak score ((58−40)/58 = 31% erosion)
4. Trust-deployment divergence: AI adoption rate (88% of organizations using AI) versus public trust in regulation (US: 31%) — a 57-percentage-point divergence

AGGI™ Q2 2026 composite reading: 58/100 (0 = perfect governance alignment; 100 = total governance failure)

Methodology: Each dimension is normalized to a 0–100 scale using its theoretical maximum and minimum. The composite is an unweighted arithmetic mean of four normalized scores. Dimension sources: (1) Stanford HAI AI Index 2026; (2) McKinsey 2026 AI Trust Maturity Survey; (3) Stanford CRFM Foundation Model Transparency Index 2025; (4) McKinsey State of AI 2025 + Stanford HAI 2026 Public Opinion data.

What a score of 58 means: The global AI ecosystem is operating at moderate-to-severe governance deficit. Enforcement capacity covers roughly one quarter of jurisdictions with active law. Corporate governance readiness covers roughly one third of organizations deploying agentic AI. Model transparency is declining at double-digit annual rates. And public trust in regulatory institutions lags AI deployment by a structurally significant margin.

The AGGI™ will be recalculated each quarter as updated primary source data becomes available. Axis Intelligence Research retains the dataset. This is the first publication of this metric; no equivalent composite existed prior to this report.

Methodology

Data collection: This report draws exclusively from primary sources — issuing organizations, institutional research bodies, and government publications. No secondary tech-media citations were used. All statistics are drawn from the most recent available publication of each source.

Coverage period: Statistics reflect data published between January 2025 and June 2026. Where data represents 2025 survey periods, this is noted explicitly in the source column.

Limitations:

• McKinsey’s AI Trust Maturity Survey (n=~500) is self-selected toward organizations with defined AI governance functions. Organizations without a governance function are under-represented.
• Gartner forecasts are projections, not observed outcomes. Forecast methodology is proprietary.
• Stanford HAI’s incident count relies on the AI Incident Database, which captures reported incidents; unreported incidents are not counted.
• The AGGI™ is Axis Intelligence’s proprietary composite and has not been independently validated.
• OECD AI Policy Observatory data is self-reported by governments and may lag actual legislative activity by 3–6 months.

Update cadence: This dataset is reviewed and updated quarterly. Next update: September 2026.

Cite This Research

Copy the embed code below to feature this research on your site. Each embed attribution links directly to this report.

<blockquote>
  <p>
    According to Axis Intelligence Research, as of Q2 2026, only 12 of 47 countries with 
    active AI legislation have enforcement mechanisms — a 74% enforcement gap. The AI 
    Governance Gap Index (AGGI™) composite score stands at 58/100.
  </p>
  <footer>
    <cite>
      <a href="https://axis-intelligence.com/ai-governance-statistics/" 
         target="_blank" rel="noopener">
        AI Governance Statistics 2026, Axis Intelligence Research
      </a>
    </cite>
  </footer>
</blockquote>

About This Dataset

License: Creative Commons Attribution 4.0 International (CC BY 4.0). You are free to share and adapt this data with attribution.

Download: A full CSV of the datasets in this report is available

Citation (APA): Axis Intelligence Research. (2026, June 7). AI governance statistics 2026. Axis Intelligence. https://axis-intelligence.com/ai-governance-statistics/

Citation (MLA): Axis Intelligence Research. “AI Governance Statistics 2026.” Axis Intelligence, 7 June 2026, axis-intelligence.com/ai-governance-statistics/.

Citation (Chicago): Axis Intelligence Research. “AI Governance Statistics 2026.” Axis Intelligence, June 7, 2026. https://axis-intelligence.com/ai-governance-statistics/.

---

Frequently Asked Questions

How many countries have AI laws in 2026?

As of early 2026, 47 countries have active AI legislation, according to the Stanford HAI 2026 AI Index Report. However, only 12 of those countries have established enforcement mechanisms — meaning 74% of AI laws have no practical enforcement structure as of mid-2026.

What is the EU AI Act enforcement date for high-risk systems?

August 2, 2026 is the primary deadline for high-risk AI system obligations under the EU AI Act. Providers of high-risk AI systems must complete conformity assessments, finalize technical documentation, and register in the EU database before placing their systems on the market. Penalties for non-compliance can reach €15 million or 3% of global annual turnover.

How much are organizations spending on AI governance?

Gartner projects spending on AI governance platforms will reach $492 million in 2026, rising to over $1 billion by 2030. This reflects the conversion of governance from an optional best practice to a regulatory compliance requirement — particularly as the EU AI Act enters full enforcement.

What is the NIST AI Risk Management Framework?

The NIST AI RMF is a voluntary framework published by the National Institute of Standards and Technology to help organizations manage AI risks across four functions: Govern, Map, Measure, and Manage. As of 2025, 33% of organizations cite NIST AI RMF as a regulatory influence on their responsible AI practices, per Stanford HAI. NIST released a Critical Infrastructure Profile concept note on April 7, 2026.

What is responsible AI maturity in organizations today?

McKinsey’s 2026 AI Trust Maturity Survey found the average organizational RAI maturity score is 2.3 out of 5 — up from 2.0 in 2025, but well below the level needed to govern autonomous AI agents effectively. Only about 30% of organizations have reached maturity level 3 or higher in strategy, governance, or agentic AI controls.

What is the Foundation Model Transparency Index?

Published by the Stanford Center for Research on Foundation Models, the FMTI scores AI companies on how openly they disclose details about their models’ training data, compute, capabilities, risks, and usage policies. The 2025 edition found average scores dropped from 58/100 in 2024 to 40/100 in 2025 — its sharpest single-year decline. IBM scored 95 (highest); xAI and Midjourney scored 14 (lowest).

Which country has the lowest trust in AI regulation?

China recorded the lowest public trust in government AI regulation at 27%, followed by the United States at 31%. EU citizens expressed the highest trust among regions covered, at 53%, per Stanford HAI 2026.

What are the main barriers to AI governance implementation?

According to Stanford HAI’s 2026 AI Index, the three main obstacles to responsible AI implementation are: gaps in knowledge (59% of organizations cite this), budget constraints (48%), and regulatory uncertainty (41%).

What is the AI Governance Gap Index (AGGI™)?

The AGGI™ is Axis Intelligence Research’s proprietary composite metric synthesizing four dimensions of the global governance gap: legislative enforcement coverage, corporate maturity deficit, model transparency erosion, and trust-deployment divergence. The Q2 2026 reading is 58/100, indicating a moderate-to-severe governance deficit. This metric is updated quarterly using verified primary source data.

How many AI incidents were recorded in 2025?

The AI Incident Database recorded 362 documented AI incidents in 2025, up 55% from 233 in 2024, according to Stanford HAI’s 2026 AI Index Report.