What are the latest funding news in the AI safety market? (September 2026)
Download our beautiful pitch about the AI safety market
In our AI safety market deck, you will find everything you need to understand the market
AI safety companies announced 12 qualifying financings worth about $435 million between April 28 and September 1, 2026.
Funding was highly concentrated, with Alice and Zenity alone raising $265 million, or about 61% of the disclosed capital.
Nine of the 12 announcements centered directly on AI agents, showing how permissions, runtime controls and governance are becoming core parts of the commercial AI safety stack.
And if you want to better understand this new industry, you can download our pitch covering the AI safety market.
Insights
- The 12 financings in this update represent about $435 million of disclosed capital, showing that AI safety is now attracting both early-stage experimentation and much larger platform investments.
- Alice and Zenity raised $265 million combined, equal to roughly 61% of all capital in this group, so headline funding remains heavily influenced by a small number of growth-stage companies.
- Eight of the 12 financings were seed rounds, yet those seed deals represented only about 23% of disclosed capital, while the four post-seed rounds captured roughly 77%.
- The average financing was roughly $36 million, while the median was only about $10.5 million, a wide gap that highlights how a few large rounds are pulling the market average upward.
- Nine of the 12 companies focus directly on AI agents, particularly permissions, tool access, runtime behavior, testing and governance, making agent safety the clearest funding theme in the AI safety market.
- AIR, Willow, Archestra, Geordie AI, CodeIntegrity, Aigentsphere and Tynapse all address what happens when autonomous agents interact with real tools and data, rather than focusing only on model behavior before deployment.
- Evaluation and red teaming remain investable categories, but Gray Swan AI and Arga Labs show how testing is moving toward realistic environments, adversarial scenarios and agent behavior rather than static model benchmarks alone.
- The market is geographically broad, with funded companies emerging from the United States, Israel, the United Kingdom, France, Australia and South Korea while generally targeting global enterprise customers.
- Valuation transparency remains limited across AI safety startups, although Alice was reported at roughly an $800 million valuation, suggesting investors are already assigning substantial value to scaled AI trust and security platforms.
As this chart shows, and as featured in our AI safety market deck, search interest in AI safety has been growing steadily
Summary table of the latest funding deals in the AI safety market as of September 2026
We define the AI safety market as products and services that reduce harms and failure modes from AI systems by measuring, mitigating, and governing model behavior and AI-specific risk.
We include AI evaluation and red teaming, safety monitoring and incident response, safety and guardrail layers, and AI trust, risk, and security management capabilities such as governance, robustness, privacy and security controls, and explainability when used for risk reduction.
We exclude generic MLOps, general cybersecurity, and general compliance or content moderation offerings unless they are specifically designed to address AI-model behavior or AI-specific threats.
You can also read our detailed analysis to understand how funding activity in the AI safety market has evolved over the last few years.
We also have a quarter-by-quarter analysis of funding activity in the market here.
Finally, you can check our complete list of fundraising deals for the AI safety market (we update this list every quarter) as well as our ranking of the most funded startups.
| Name | When | Amount in $ | Round Type | Category |
|---|---|---|---|---|
| Tynapse | April 28, 2026 | $3.17M | Seed | Runtime AI guardrails & agent trust infrastructure |
| Aigentsphere | April 29, 2026 | $4M | Seed | AI-agent governance & inventory management |
| White Circle | May 12, 2026 | $11M | Seed | Production AI safety monitoring & guardrails |
| CodeIntegrity | May 27, 2026 | $5M | Seed | Runtime guardrails & AI-agent application security |
| Gray Swan AI | May 28, 2026 | $40M | Series A | AI evaluation, adversarial testing & red teaming |
| Geordie AI | May 28, 2026 | $30M | Series A | AI-agent security, governance & runtime control |
| Archestra | June 2, 2026 | $10M | Seed | AI-agent security infrastructure & policy enforcement |
| Willow | June 4, 2026 | $7M | Seed | AI-agent identity, access governance & oversight |
| Zenity | August 3, 2026 | $125M | Series C | AI-agent security, monitoring & governance |
| Alice | August 25, 2026 | $140M | Growth financing | AI trust, adversarial testing & security |
| Arga Labs | August 26, 2026 | $10M | Seed | AI-agent evaluation, simulation & safety testing |
| AIR | September 1, 2026 | $50M total | Two seed rounds | AI-agent security, add-on vetting & runtime governance |
All the latest funding deals during in the AI safety market as of September 2026
AIR announced $50 million across two seed financings in September 2026 to secure the growing software ecosystem around AI agents.
When was it?
AIR publicly announced the financings on September 1, 2026.
Who are they?
AIR discovers enterprise AI agents, vets the skills, plugins and MCP components they use, and blocks unsafe interactions before they reach an agent.
Geographical focus?
AIR is based in the United States and is targeting enterprises deploying AI agents across global operations.
Why do we include them in the AI safety market?
AIR belongs in the AI safety market because its core product reduces AI-specific supply-chain, prompt-injection, permission and runtime risks created by autonomous agents.
What is the company stage?
AIR is at the seed and early-commercial stage, with more than 20 corporate customers reported when the company emerged from stealth.
How much did they raise?
AIR raised $50 million in total across the two financings.
What round is it?
AIR raised an initial $10 million seed led by Sequoia followed by a $40 million seed financing led by Greenoaks.
Why did they raise?
AIR raised the capital to build out its agent-security platform, expand enterprise deployments and provide broader protection across the emerging AI add-on ecosystem.
Arga Labs raised $10 million in August 2026 to build realistic simulation environments for testing and training enterprise AI agents.
When was it?
Arga Labs announced the financing on August 26, 2026.
Who are they?
Arga Labs builds stateful digital twins of enterprise software so AI agents can be tested and trained without touching real production systems.
Geographical focus?
Arga Labs is a United States company building infrastructure for enterprises deploying agents across widely used business software.
Why do we include them in the AI safety market?
Arga Labs belongs in the AI safety market because realistic pre-deployment testing can expose unsafe agent behavior, permission problems and workflow failures before agents reach production.
What is the company stage?
Arga Labs is at the seed and early-commercial stage, with a working platform focused on enterprise agent testing and training.
How much did they raise?
Arga Labs raised $10 million in this round.
What round is it?
Arga Labs raised a seed round led by General Catalyst.
Why did they raise?
Arga Labs raised the capital to expand its simulation environments, engineering capabilities and coverage of enterprise software used by AI agents.
This chart, featured in our AI safety market deck, compares the main business model options for AI alignment research labs
Alice raised $140 million in August 2026 to expand AI adversarial testing, guardrails and security for frontier labs and enterprises.
When was it?
Alice announced the financing on August 25, 2026.
Who are they?
Alice, previously known as ActiveFence, tests AI systems against real-world attacks and provides defenses against jailbreaks, prompt injection and other adversarial behavior.
Geographical focus?
Alice is Israeli-founded and serves major AI laboratories and enterprise customers globally.
Why do we include them in the AI safety market?
Alice belongs in the AI safety market because adversarial testing, model guardrails and protection against AI-specific attacks are central parts of its platform.
What is the company stage?
Alice is a growth-stage company approaching $100 million in annual recurring revenue and working with most of the leading frontier AI laboratories.
How much did they raise?
Alice raised $140 million in this financing, bringing reported total funding to $280 million.
What round is it?
Alice completed a growth financing led by Apax Digital.
Why did they raise?
Alice raised the capital to advance its AI security platform, expand its adversarial-intelligence capabilities and scale commercial operations.
Zenity raised $125 million in August 2026 to scale security and governance infrastructure for autonomous enterprise AI agents.
When was it?
Zenity announced the financing on August 3, 2026.
Who are they?
Zenity discovers enterprise AI agents, analyzes their permissions and behavior, and applies controls before risky actions affect business systems.
Geographical focus?
Zenity is Israeli-founded and serves large enterprises globally, with growing activity across North America, Europe and Asia-Pacific.
Why do we include them in the AI safety market?
Zenity belongs in the AI safety market because its platform monitors autonomous behavior and enforces security and governance rules around what AI agents can access and do.
What is the company stage?
Zenity is at the growth stage, with a mature enterprise platform, international customers and approximately $185 million in reported total funding.
How much did they raise?
Zenity raised $125 million in this round.
What round is it?
Zenity's financing was a Series C led by Norwest Venture Partners.
Why did they raise?
Zenity raised the capital to expand globally, grow its product and research capabilities and support enterprise adoption of autonomous AI agents.
This chart, featured in our AI safety market deck, shows how HiddenLayer is positioned in AI safety
Willow raised $7 million in June 2026 to give enterprises identity, access and governance controls for autonomous workplace agents.
When was it?
Willow announced the financing on June 4, 2026.
Who are they?
Willow gives each enterprise AI agent an identity, scoped access to approved tools, runtime controls and an audit trail tied to a human user.
Geographical focus?
Willow is based in Herzliya, Israel, and targets enterprises deploying AI agents across global operations.
Why do we include them in the AI safety market?
Willow belongs in the AI safety market because its product reduces unsafe agent autonomy through permissions, access governance, runtime guardrails and auditable activity records.
What is the company stage?
Willow is at the seed and early-commercial stage, with its platform already used in production by thousands of employees at Wix.
How much did they raise?
Willow raised $7 million in this round.
What round is it?
Willow's financing was a seed round led by Hetz Ventures.
Why did they raise?
Willow raised the capital to accelerate product development and expand go-to-market activity as enterprises deploy more autonomous agents.
Archestra raised $10 million in June 2026 to secure how enterprise AI agents access sensitive data, tools and credentials.
When was it?
Archestra announced the financing on June 2, 2026.
Who are they?
Archestra sits between AI agents and enterprise systems, checking identity and policy before allowing agents to access data or execute connected tools.
Geographical focus?
Archestra is based in London and serves international enterprises, including reported Fortune 500 customers.
Why do we include them in the AI safety market?
Archestra belongs in the AI safety market because its platform addresses AI-specific risks such as prompt injection, data exfiltration and uncontrolled agent access.
What is the company stage?
Archestra is at the seed and early product-market-fit stage, with its platform already running in production at several Fortune 500 companies.
How much did they raise?
Archestra raised $10 million in this round.
What round is it?
Archestra's financing was a seed round led by 20VC.
Why did they raise?
Archestra raised the capital to grow engineering and go-to-market teams, expand enterprise deployments and develop its open-source ecosystem.
In our AI safety market deck, we identify pain points entrepreneurs should prioritize
Gray Swan AI raised $40 million in May 2026 to expand adversarial testing and security evaluations for advanced AI systems.
When was it?
Gray Swan AI announced the financing on May 28, 2026.
Who are they?
Gray Swan AI pressure-tests models and AI applications to uncover jailbreaks, exploitable weaknesses, unsafe outputs and other adversarial failure modes.
Geographical focus?
Gray Swan AI emerged from Carnegie Mellon University's Pittsburgh research ecosystem and works with frontier AI laboratories and enterprise customers globally.
Why do we include them in the AI safety market?
Gray Swan AI belongs in the AI safety market because AI evaluation, adversarial testing and red teaming are the company's core products.
What is the company stage?
Gray Swan AI is at the Series A and early-growth stage, with major frontier laboratories already using its evaluation capabilities.
How much did they raise?
Gray Swan AI raised $40 million in this round.
What round is it?
Gray Swan AI's financing was a Series A co-led by Wing Venture Capital and Madrona.
Why did they raise?
Gray Swan AI raised the capital to scale its enterprise security product, grow its team and deepen its work with major AI laboratories.
Geordie AI raised $30 million in May 2026 to expand security and governance infrastructure for enterprise AI agents.
When was it?
Geordie AI announced the financing on May 28, 2026.
Who are they?
Geordie AI gives enterprises visibility into autonomous agents and controls their permissions, access, behavior and interactions with business systems.
Geographical focus?
Geordie AI is headquartered in London and is expanding its commercial presence in the United States while serving global enterprises.
Why do we include them in the AI safety market?
Geordie AI belongs in the AI safety market because its core platform is purpose-built to secure and govern autonomous AI agents in production.
What is the company stage?
Geordie AI is at the Series A and early-growth stage, with strong enterprise adoption and rapid recurring-revenue growth reported around the financing.
How much did they raise?
Geordie AI raised $30 million in this round.
What round is it?
Geordie AI's financing was a Series A led by Balderton Capital.
Why did they raise?
Geordie AI raised the capital to enhance its security platform, expand engineering and build a larger United States go-to-market operation.
This market map, featured in our AI safety market deck, highlights top companies and startups in the AI safety market
CodeIntegrity raised about $5 million in May 2026 to place deterministic runtime controls around unpredictable enterprise AI agents.
When was it?
CodeIntegrity announced the financing on May 27, 2026.
Who are they?
CodeIntegrity checks AI-agent tool calls, permissions and data flows before execution so risky actions can be blocked before reaching production systems.
Geographical focus?
CodeIntegrity is based in San Francisco and initially targets enterprises deploying agentic applications in sensitive production environments.
Why do we include them in the AI safety market?
CodeIntegrity belongs in the AI safety market because its core product addresses prompt injection, data leakage, excessive autonomy and other AI-agent-specific runtime risks.
What is the company stage?
CodeIntegrity is at the seed and early-commercial stage, with pilots underway in regulated industries when the round was announced.
How much did they raise?
CodeIntegrity announced a $5 million seed round, with some databases reporting the unrounded amount as approximately $4.8 million.
What round is it?
CodeIntegrity's financing was a seed round led by SYN Ventures.
Why did they raise?
CodeIntegrity raised the capital to expand its runtime control platform and help enterprises deploy autonomous agents with enforceable security policies.
White Circle raised $11 million in May 2026 to monitor, secure and control the behavior of AI systems running in production.
When was it?
White Circle announced the financing on May 12, 2026.
Who are they?
White Circle monitors live AI applications and applies controls around model outputs, user behavior, tool calls, hallucinations and malicious activity.
Geographical focus?
White Circle is based in Paris and is expanding across Europe, the United Kingdom and the United States.
Why do we include them in the AI safety market?
White Circle belongs in the AI safety market because its platform detects AI-specific failures such as jailbreaks, prompt injection, drift and unsafe model behavior and can intervene in real time.
What is the company stage?
White Circle is at the seed and early product-market-fit stage, with more than one billion API requests processed and several large customers reported around the financing.
How much did they raise?
White Circle raised $11 million in this round.
What round is it?
White Circle's financing was a seed round backed by a group of prominent AI and technology executives.
Why did they raise?
White Circle raised the capital to accelerate product development, hire across major markets and expand its enterprise customer base.
This chart, featured in our AI safety market deck, shows annual funding in AI safety startups
Aigentsphere raised $4 million in April 2026 to help enterprises monitor, govern and control expanding fleets of autonomous AI agents.
When was it?
Aigentsphere's financing was publicly reported on April 29, 2026.
Who are they?
Aigentsphere provides a central system for registering AI agents, monitoring their performance and costs, enforcing policies and generating governance reports.
Geographical focus?
Aigentsphere is based in Sydney and is expanding operations across Australia and the United States.
Why do we include them in the AI safety market?
Aigentsphere belongs in the AI safety market because its core platform addresses agent sprawl, accountability, policy enforcement and operational oversight for autonomous AI systems.
What is the company stage?
Aigentsphere is at the seed and early-commercial stage, with enterprise pilots and an initial multi-year customer contract reported around the financing.
How much did they raise?
Aigentsphere raised $4 million in the announced seed financing.
What round is it?
Aigentsphere's financing was a seed round led by Main Sequence.
Why did they raise?
Aigentsphere raised the capital to grow its engineering team, accelerate enterprise deployments and expand its presence in Australia and the United States.
Tynapse raised approximately $3.17 million in April 2026 to commercialize runtime trust and security infrastructure for AI agents.
When was it?
Tynapse announced the seed financing on April 28, 2026.
Who are they?
Tynapse validates AI-agent responses and actions during execution and can block hallucinations, data leakage, jailbreaks and other unsafe behavior.
Geographical focus?
Tynapse is based in South Korea and is developing its AI trust infrastructure for Korean and international enterprise deployments.
Why do we include them in the AI safety market?
Tynapse belongs in the AI safety market because its core product is specifically designed to detect and control model and agent failure modes at runtime.
What is the company stage?
Tynapse is at the seed and early-product stage, with a focused AI trust layer and initial technical and commercial validation.
How much did they raise?
Tynapse raised KRW 4.5 billion, reported at approximately $3.17 million in the funding dataset used for this analysis.
What round is it?
Tynapse's financing was a seed round led by Mirae Asset Venture Investment.
Why did they raise?
Tynapse raised the capital to develop and commercialize infrastructure that makes autonomous AI-agent execution safer, more reliable and easier to audit.
Related blog posts
- The evolution of funding activity in the AI safety market
- The main fundraising trends in the AI safety market
- How strong is fundraising in the AI safety market right now?
