AI Governance Funding: A Quarter-by-Quarter Analysis

All funding deals in the AI governance market over the last years

Below is the table listing all the deals. You can find our methodology at the end of this page.

If you want a deeper understanding of the market and its current dynamics, get our report covering the AI Governance Market.

Is AI governance funding actually taking off now?

Q2 2026 is the first quarter where AI governance funding looks like a real market-wide acceleration rather than another spike caused by one oversized round.

We tracked 27 financings worth $462.8 million in Q2 2026, up from only 5 deals and $97.8 million in Q1. More importantly, the median round rose from $6 million to $10 million while capital excluding rounds above $50 million jumped from $39.8 million to $338.8 million. Those three things had never moved up together during the previous four quarters.

The path to that point was messy. Q3 2025 produced a large headline total but leaned heavily on Noma Security and Irregular. Q4 then collapsed in volume. Q1 2026 looked like a rebound until we stripped out WitnessAI's $58 million round. Q2 is where the shape of the market genuinely changed.

We would still call this an emerging acceleration rather than a settled multi-quarter trend. One quarter can mark a turning point, but another strong quarter would make the case much harder to dismiss.

Quarter	Deals	Capital	Median round	Capital excluding >$50M
Q2 2025	11	$130.8M	$7.0M	$130.8M
Q3 2025	15	$253.4M	$6.5M	$153.4M
Q4 2025	6	$62.5M	$11.25M	$62.5M
Q1 2026	5	$97.8M	$6.0M	$39.8M
Q2 2026	27	$462.8M	$10.0M	$338.8M

Was the Q2 2026 AI governance boom really broad, or did two big rounds inflate it?

The Q2 2026 AI governance boom was broad enough that removing the biggest rounds barely changes the story.

Straiker raised $64 million and Arcade.dev raised $60 million, yet the largest deal represented only 13.8% of quarterly funding. Compare that with Q1 2026, when WitnessAI alone accounted for 59.3%.

The top three Q2 deals captured 37.6% of funding, down from 96.1% in Q1. Even the top five represented barely more than half of the quarter. That is unusually wide distribution for a young funding category.

The middle of the market filled out too. Fifteen companies raised between $5 million and $20 million, while another five raised between $20 million and $50 million. The boom therefore reached companies well beyond the handful already capable of raising $50 million-plus rounds.

Quarter	Largest deal share	Top 3 share	Top 5 share	Deals above $50M
Q2 2025	38.2%	66.9%	84.2%	0
Q3 2025	39.5%	71.0%	84.2%	1
Q4 2025	28.8%	80.0%	98.4%	0
Q1 2026	59.3%	96.1%	100%	1
Q2 2026	13.8%	37.6%	52.7%	2

Why did AI governance funding look so strong in Q3 2025?

Q3 2025 was a genuine improvement in AI governance deal activity, but the headline funding number made the quarter look hotter than it really was.

Total funding almost doubled from $130.8 million to $253.4 million, while deal count rose from 11 to 15. The median round actually slipped from $7 million to $6.5 million.

The explanation sits at the top of the distribution. Noma Security raised $100 million, while Irregular announced a $30 million Seed and an approximately $50 million Series A that had closed only weeks apart. Those deals pulled the quarterly total upward much faster than the rest of the market.

Once rounds above $50 million are removed, capital grew by only 17.3%. Q3 was still important because more companies were raising, but the market had not yet developed the depth we see later.

What happened to AI governance funding in Q4 2025?

AI governance funding simply ran out of deal volume in Q4 2025, even though investors were still willing to finance the companies that did make it to market.

The number of deals fell from 15 to 6 and capital dropped 75.3% to $62.5 million. Capital excluding the biggest rounds also fell sharply, so there is no large-deal distortion hiding the contraction.

Yet the median financing climbed to $11.25 million, its highest point in the five-quarter period. The six companies that raised were not scraping together tiny rounds.

Q4 looks like a thin pipeline quarter. With so few financings, a handful of delayed or accelerated announcements could move the totals dramatically. That sort of volatility is normal when a funding category still lacks much depth.

Was Q1 2026 really an AI governance funding recovery?

Q1 2026 looked like a recovery in total dollars, but the underlying AI governance market was weaker than Q4 on several measures.

Capital increased by 56.5%, yet there were only five deals and the median round fell from $11.25 million to $6 million. Capital excluding rounds above $50 million dropped another 36.3%.

WitnessAI's $58 million financing did most of the work. Add Fiddler AI's $30 million Series C and the two companies accounted for roughly 90% of all money raised during the quarter.

There was still something useful happening underneath. Both companies attracted strategic investors and were selling broader control, monitoring and governance infrastructure for enterprise AI. Q1 showed that established governance vendors could raise serious follow-on capital. It just did not show a broad funding recovery.

Are more AI governance startups getting funded, or are investors simply writing bigger checks?

Both things happened in Q2 2026: more AI governance companies raised money, and the typical new financing became meaningfully larger.

Twelve first financings appeared during the quarter, the highest number we found in the five-quarter period. Those new entrants raised $104.2 million, with a median first financing of $8.4 million. In Q1, the median was only $1.9 million.

At the same time, follow-on rounds remained the larger pool of capital. Fifteen follow-ons attracted $358.6 million, showing that investors were also pushing more money into companies that had already survived an earlier funding round.

The round-size distribution tells the same story from another angle. Q2 had only four financings below $5 million, compared with 20 between $5 million and $50 million. AI governance now has a much more visible middle layer of institutional Seed and Series A companies than it did a year earlier.

Quarter	First financings	Follow-ons	Median first financing	Median follow-on
Q2 2025	6	5	$4.7M	$14.5M
Q3 2025	9	6	$4.7M	$14.7M
Q4 2025	3	3	$4.0M	$15.0M
Q1 2026	2	3	$1.9M	$30.0M
Q2 2026	12	15	$8.4M	$10.0M

Which part of the AI governance market has the strongest funding momentum?

AI Policy Enforcement is still the clearest core of the AI governance funding market.

The category accounted for roughly one-third to almost half of all deals in every quarter we studied: 45.5% in Q2 2025, 33.3% in Q3, 33.3% in Q4, 40% in Q1 2026 and 44.4% in Q2.

That consistency is more useful than one spectacular funding quarter. Policy Enforcement includes companies controlling AI access, agent actions, runtime behavior and enterprise policies, putting it close to the point where AI systems actually touch company data and tools.

The latest group includes businesses such as Straiker, Arcade.dev, Archestra, CodeIntegrity, Tenet Security and ZeroDrift. The product descriptions vary, but investors are repeatedly funding the same basic problem: enterprises want agents to do useful work without giving them unrestricted freedom inside production systems.

Q2 2026 category	Deals	Capital	Current read
AI Policy Enforcement	12	$200.0M	Established core
AI Governance Platforms	8	$113.8M	Deepening quickly
Governance Evidence Tools	2	$56.5M	Small but well funded
AI Assurance Services	2	$50.0M	Reappearing
AI Audit Software	2	$38.0M	Strong ticket sizes
AI Compliance Tools	1	$4.5M	Still peripheral

Are AI assurance, audit and evidence tools actually breaking out?

AI assurance, audit and evidence tools are attracting much bigger checks now, although their quarter-to-quarter activity is still too uneven to call them fully established funding categories.

AI Assurance Services surged to $97.1 million in Q3 2025, largely because Irregular raised two rounds worth about $80 million combined. The category then went quiet for two quarters before Gray Swan and another company brought it back to $50 million in Q2 2026.

Audit Software tells a slightly different story. It returned in Q2 with only two deals, but their median size reached $19 million. Coval's $28 million Series A was particularly notable because the company is testing and simulating autonomous voice and chat agents for enterprise customers.

Governance Evidence Tools remain tiny in deal count, yet Patronus AI's $50 million Series B pushed the category to $56.5 million in the latest quarter. Patronus is also moving beyond conventional model evaluation into simulation environments for training and testing agents.

The larger ticket sizes are worth taking seriously. What is still missing is regular deal flow across several consecutive quarters.

Are AI Governance Platforms becoming a second major funding category?

AI Governance Platforms now look like the strongest second category behind Policy Enforcement.

Eight Governance Platform companies raised $113.8 million in Q2 2026, with a median financing of $11.25 million. That is a much more convincing base than Q1, when the category represented 60% of all deals simply because the whole quarter contained only five financings.

These companies are building inventories, control planes, policy systems and audit layers that sit above individual AI applications. Fiddler AI describes its product as a control plane for compound AI, while Willow focuses on identities, scoped permissions and audit trails for enterprise agents. Aigentsphere is building a system of record for agents.

The common idea is becoming easier to see: companies need one place to know which AI systems exist, who controls them, what they can access and what they have done. Governance software is turning into a distinct infrastructure layer.

Are AI governance investors moving toward later-stage companies?

AI governance funding is still overwhelmingly an early-stage market, despite the brief later-stage concentration we saw in Q1 2026.

Q1 had WitnessAI at Series B and Fiddler AI at Series C, so later-stage deals made up 40% of transactions and almost 90% of capital. With only five rounds in the quarter, two mature companies were enough to create that impression.

Q2 swung straight back toward earlier stages. We counted 18 Seed rounds and 7 Series A rounds, meaning 92.6% of financings came from those two stages. Together they also captured about 87% of the money.

We also checked whether investors were consistently moving toward younger companies, but the founding-year field was too unreliable to answer that properly. Forty-six of 64 records showed 2009, including obviously newer startups. After removing those placeholder values, only 16 usable observations remained. Better to leave the age question open than manufacture a clean trend from bad data.

Is AI governance capital going mostly to companies that have already raised?

Follow-on rounds still take most of the AI governance money, but there is plenty of room for new companies to get funded.

In Q2 2026, follow-ons represented 15 of 27 deals and absorbed roughly three-quarters of quarterly capital. Companies such as Straiker, Arcade.dev, Gray Swan, Coval and Patronus had already raised before coming back for larger rounds.

At the same time, the first-financing pipeline became much healthier. As seen above, Q2 produced a record 12 first financings in our five-quarter window, and the typical first round was far larger than in Q1.

That mix makes sense for a market entering its next phase. Investors are still funding brand-new approaches while putting much bigger amounts behind companies that have already found customers or shown that their technology works in production.

Are the same VCs dominating AI governance funding every quarter?

AI governance funding is drawing in a wide mix of investors rather than being controlled by the same small circle of venture firms.

We found 33 distinct disclosed lead investors across the 27 Q2 2026 transactions because several deals had more than one lead. SYN Ventures appeared twice; almost everyone else appeared once.

Across the full five-quarter period, recurring leads were surprisingly rare. Insight Partners was the only lead we found appearing in two different quarters, Q3 and Q4 2025.

Quarterly fundraising is too lumpy to conclude that a fund has left the market just because it disappears for a few months. Still, the current investor map is broad. AI governance has moved beyond a niche financed by two or three specialist firms.

Why are corporate investors suddenly showing up in AI governance funding?

Corporate investors are moving into AI governance because agent security and control now touch systems those companies already sell, operate or depend on.

Strategic participation was rare through most of 2025. It then appeared in 40% of Q1 2026 deals and 37% of Q2 deals. The Q1 figure came from only two transactions, so Q2 gives us much better evidence.

Morgan Stanley and Wipro invested in Arcade.dev, which controls authentication and permissions for agents taking actions inside enterprise systems. Citi Ventures and Workday Ventures backed Straiker. Snowflake Ventures invested in Gray Swan. Patronus AI brought in Datadog and Samsung.

What makes this interesting is the stage mix. Nine of the ten Q2 transactions with a strategic investor were Seed or Series A. Corporate money is arriving before companies become mature, suggesting these investors want early exposure to infrastructure that may sit directly next to their own AI products and customers.

Quarter	Deals with strategic investors	Share of deals	What we see
Q2 2025	1	9.1%	Occasional
Q3 2025	1	6.7%	Still rare
Q4 2025	1	16.7%	Starting to appear
Q1 2026	2	40.0%	High share, tiny quarter
Q2 2026	10	37.0%	Broad across Seed and Series A

Is AI governance funding becoming less US-centric?

AI governance deal activity is spreading well beyond North America, while the biggest checks still lean heavily toward North American companies.

North America accounted for 90.9% of deals in Q2 2025. By Q2 2026, its share had fallen to 55.6%. Europe, the Middle East and Asia-Pacific each produced four financings in the latest quarter.

The capital split moved much less. North American companies still captured 75.1% of Q2 funding. Europe received 15.3%, while the Middle East and Asia-Pacific remained in single digits.

The startup base is internationalizing faster than the money. Israel is producing several runtime-security companies, Europe is showing more activity in governance and policy infrastructure, and South Korea is beginning to appear in AI safety and enforcement. The largest institutional rounds, however, are still mostly going to US-based companies.

What is actually driving the AI governance funding boom?

The current AI governance funding boom is closely tied to companies moving AI agents from demos into production, where permissions, runtime behavior and failure modes suddenly become real operational problems.

Arcade.dev's funding announcement gives a clear version of that story. Its product checks what an agent is allowed to do at the moment it acts, then records those actions for audit. Archestra says multiple Fortune 500 companies are already running its infrastructure in production and frames the bottleneck around connecting agents to corporate data without losing control.

Straiker combines agent discovery, adversarial testing and runtime protection. Its funding followed rapid enterprise adoption and a claimed 15-fold increase in run-rate revenue in less than a year. Willow came out of stealth after running inside Wix, where it says roughly 5,000 people were already using the platform weekly.

Evaluation companies are seeing the same shift from another direction. Coval simulates voice and chat agents before companies deploy them at scale. Patronus is building digital environments where agents can be trained and tested against more complex situations.

These companies are solving different pieces of the same production headache: agents are being given tools, permissions and autonomy faster than the control infrastructure around them has matured.

Is regulation also pushing money into AI governance?

Regulation is helping AI governance demand, especially in Europe, but we would not treat it as the main explanation for the funding surge.

EU rules for providers of general-purpose AI models began applying in August 2025, covering documentation, copyright, transparency and additional safety requirements for models with systemic risk. The European Commission's enforcement powers then became applicable in August 2026.

That creates obvious work for governance teams, particularly around evidence, model documentation, risk management and auditability. European startups such as AI Score and Velatir are explicitly pitching continuous governance rather than occasional compliance exercises.

Still, the geography of the funding boom points elsewhere too. North America took three-quarters of Q2 capital, and many of the largest rounds went to companies focused on agent permissions, security, simulation and production control. Regulation adds demand, but enterprise deployment is currently the broader catalyst.

Does the Q2 2026 AI governance breakout still look real as of now?

The early evidence since Q2 suggests that investor interest in AI governance has stayed hot rather than disappearing immediately after one exceptional quarter.

We keep the quarterly statistics in this article strictly capped at June 30, 2026, so later deals do not contaminate the five-quarter comparison. But the post-period financings are useful as a check on the direction of travel.

Zenity subsequently raised a $125 million Series C to expand its AI-agent security platform. AIR Security emerged with $50 million raised across two Seed rounds focused on vetting agent skills, plugins and MCP servers. AI Score followed with a $5.4 million round for continuous enterprise governance, while Velatir announced €5 million for its own governance layer.

The sizes and stages are all over the map, which is actually useful here. We are seeing fresh Seed companies, follow-on rounds and a major later-stage financing around the same general problem. Q2's breakout therefore has some early follow-through, even though these later financings sit outside our formal analysis window.

What could still derail the AI governance funding story?

The biggest risk in AI governance funding is that Q2's breadth turns out to be a burst of category formation that produces far more vendors than durable winners.

There are already several companies pitching agent inventories, runtime controls, permissions, red teaming, simulation, audit evidence and governance control planes. Those products overlap more than their category labels sometimes suggest. A large enterprise probably will not want a separate platform for every one of those functions.

The market also remains young. Policy Enforcement has held up across several quarters, but Assurance, Audit and Evidence still move in bursts. Geographic deal activity is broadening faster than large-ticket funding. Later-stage depth remains limited.

Then there is platform risk. Cloud providers, model companies, identity vendors, cybersecurity incumbents and observability companies can all add governance features to products customers already use. Independent vendors will need to prove they own a genuinely difficult control point rather than a temporary gap in the stack.

The funding market looks healthier these days, but the next test is commercial consolidation: which of these companies can turn an urgent technical problem into a product enterprises keep paying for after the first wave of agent deployment?

What is the AI governance funding trend actually telling us?

AI governance funding has moved from a thin and erratic market into a much deeper one.

The clearest change is breadth. More companies are raising, mid-sized rounds are far more common, and large deals no longer explain most of the quarterly total.

Policy Enforcement has the best evidence of durable demand. Governance Platforms are developing into a second serious category, while Audit, Assurance and Evidence have started attracting much larger checks without yet showing the same consistency.

The market is still early-stage at its core. Seed and Series A dominate deal count, and new companies continue to appear even as the strongest existing startups raise much larger follow-ons.

Investor participation is widening too. We see many different lead investors, increasing corporate participation and more activity outside North America, although the biggest pools of capital remain concentrated in the US.

The timing also makes commercial sense. Enterprises are giving agents access to code, databases, business applications and customer interactions. That pushes governance from a policy document into infrastructure that has to work in real time.

For now, focusing only on the jump in quarterly dollars misses the more interesting change. Many different investors are now funding many different companies to solve increasingly concrete production problems around AI agents.