All the fundraising deals in the AI safety market (from Q1 2025 to Q1 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
The AI safety market raised over $817 million across 33 verified deals between Q1 2025 and Q1 2026, spanning red teaming, guardrails, governance, and AI security platforms.
Most rounds were early-stage, but a handful of mega-deals reshaped the total, with five companies alone accounting for more than half of all capital raised during the period.
The shift from protecting language models to securing AI agents became the defining theme of the market by late 2025 and into 2026.
And if you want to better understand this new industry, you can download our pitch covering the AI safety market.
In our AI safety market deck, we will give you useful market maps and grids
Summary table of the funding deals in the AI safety market (last 5 quarters)
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/guardrail layers, and AI trust-risk-security management capabilities such as governance, robustness, privacy/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.
Also, you should know that we have a dedicated page, updated weekly, with all the latest fundraising deals in the AI safety market.
| Name | What they do | Amount | Quarter | Source(s) |
|---|---|---|---|---|
| Singulr AI | Enterprise AI governance platform that detects shadow AI, data leakage, and AI sprawl in regulated companies. | $10.0M | Q1 2025 | BusinessWire |
| Dreadnode | Offensive AI security tools that stress-test AI systems for prompt injection and model bypass vulnerabilities. | $14.0M | Q1 2025 | Dreadnode Blog |
| AIceberg | AI firewall that validates AI traffic across text, speech, images, and code to enforce safety and compliance policies. | $10.0M | Q1 2025 | PR Newswire |
| GetReal Security | Detects and mitigates AI-generated threats like deepfakes and impersonation attacks for governments and enterprises. | $17.5M | Q1 2025 | GetReal, TechCrunch |
| SplxAI | Automates offensive security testing for chatbots and agentic AI systems to find prompt injection and context leakage. | $7.0M | Q1 2025 | SplxAI Blog, BusinessWire |
| Straiker | Secures enterprise AI applications and agents with continuous red-team simulations and runtime threat blocking. | $21.0M | Q1 2025 | Straiker Blog, PR Newswire |
| Yrikka AI | Automates AI red teaming so validation cycles shrink from months to minutes for critical AI systems. | $1.5M | Q2 2025 | SiliconAngle |
| Aurascape | AI-native security platform that provides visibility, policy control, and governance over enterprise AI usage and shadow AI. | $50.0M | Q2 2025 | Aurascape, SecurityWeek |
| Virtue AI | AI safety platform combining automated red teaming, guardrails, and security workflows for enterprise AI deployments. | $30.0M | Q2 2025 | BusinessWire |
| Pillar Security | Gives enterprises oversight, detection, testing, and response capabilities for AI software and AI applications. | $9.0M | Q2 2025 | Pillar Blog, SecurityWeek |
| Trustible | AI governance platform that helps companies identify, measure, mitigate, and document AI risk before production deployment. | $4.6M | Q2 2025 | PR Newswire |
| Repello AI | Runs automated GenAI red teaming and recommends targeted mitigations and guardrails for text, image, and audio workflows. | $1.225M | Q2 2025 | Repello Blog |
| Modulos | Automates AI governance and compliance workflows for frameworks like the EU AI Act, NIST AI RMF, and ISO 42001. | $10.9M | Q3 2025 | Modulos Press |
| Promptfoo | Open-source AI security testing and red-teaming platform that helps developers find prompt injection and LLM risks. | $18.4M | Q3 2025 | Promptfoo Blog, SecurityWeek |
| Noma Security | End-to-end AI and agent security platform covering discovery, governance, posture management, and runtime protection. | $100.0M | Q3 2025 | Noma Blog, PR Newswire |
| AIM Intelligence | Builds AI security and safety products including automated red teaming, guardrails, and real-time supervision for AI services. | $1.3M | Q3 2025 | EIN Presswire |
| Airia | Enterprise AI security, governance, and orchestration layer designed for safe AI deployment at scale. | $100.0M | Q3 2025 | Airia, SiliconAngle |
| Keycard | Identity and access infrastructure built specifically for AI agents, with task-scoped access controls instead of broad standing permissions. | $38.0M | Q4 2025 | SecurityWeek, SiliconAngle |
| Darwin AI | Builds governance and workflow tools for governments to deploy AI safely, transparently, and at scale. | $15.0M | Q4 2025 | Darwin AI, PR Newswire |
| Truth Systems | Turns company policies into enforceable AI guardrails, especially for legal and regulated workflows. | $4.0M | Q4 2025 | Artificial Lawyer |
| Polygraf AI | Uses proprietary small language models to identify AI-driven risks, prevent fraud, and protect enterprise intelligence in defense contexts. | $9.5M | Q4 2025 | Polygraf, VentureBeat |
| Portal26 | Helps enterprises govern GenAI adoption with shadow-AI detection, policy enforcement, risk management, and forensic audit capabilities. | $9.0M | Q4 2025 | Portal26, SecurityWeek |
| Haize Labs | Stress-tests and red-teams AI systems for trust, safety, and reliability before production deployment. | $12.5M | Q4 2025 | Signalbase |
| AI Score | Centralized AI governance and management platform that gives companies real-time visibility and control over AI usage. | $1.0M | Q4 2025 | EU-Startups |
| Runlayer | Secures the Model Context Protocol (MCP) so AI agents can interact with enterprise systems safely and with governance controls. | $11.0M | Q4 2025 | TechCrunch |
| Vijil | Provides trust infrastructure for AI agents, including testing, runtime governance, telemetry, and continuous hardening. | $17.0M | Q4 2025 | BusinessWire, SiliconAngle |
| Lumia Security | Gives enterprises network-level visibility and control over AI and autonomous-agent usage, especially in financial services. | $18.0M | Q4 2025 | Lumia Blog |
| Ciphero | AI verification layer that captures, verifies, and governs enterprise AI interactions, including shadow AI and agentic workflows. | $2.5M | Q4 2025 | Ciphero Blog, SecurityWeek |
| LMArena | Head-to-head evaluation platform that benchmarks model behavior, quality, and safety using large-scale human preference data. | $150.0M | Q1 2026 | TechCrunch, PR Newswire |
| WitnessAI | Secures and governs enterprise AI and agents by controlling data flows and enforcing policy requirements across agentic workflows. | $58.0M | Q1 2026 | PR Newswire |
| SecStrike | Thailand-based offensive security and AI research firm building automated red-teaming tools for AI systems. | $1.25M | Q1 2026 | Nation Thailand |
| Fiddler AI | Provides AI observability, evaluation, monitoring, policy controls, and governance as a control plane for AI agents. | $30.0M | Q1 2026 | Fiddler Press, BusinessWire |
| JetStream Security | Gives enterprises real-time visibility, governance, and control for AI agents, bots, models, and MCP-connected systems. | $34.0M | Q1 2026 | JetStream, SecurityWeek |
In our AI safety market deck, we identify pain points entrepreneurs should prioritize
How has funding activity in the AI safety market changed over time?
Q1 2026 was the most capital-intensive quarter, pulling in $273M across just 5 deals, but that figure is heavily shaped by LMArena's $150M Series A, which alone represents more than half of the quarter's total.
Q1 2025 was the quietest quarter, with $79.5M raised across 6 deals and no round exceeding $22M, reflecting the market's earlier stage before large platform plays emerged.
From Q4 2025 to Q1 2026, total AI safety market funding jumped roughly 99%, and compared to Q1 2025 (one year earlier), the same quarter in 2026 raised about 244% more.
If you remove the top two deals per quarter (which tend to skew results significantly), the remaining deal activity in the AI safety market looks more stable, averaging roughly $10M to $15M per round, with a steady but gradual increase over time rather than dramatic acceleration.
| Quarter | Number of deals | Total raised | Comment |
|---|---|---|---|
| Q1 2025 | 6 | $79.5M | Market warm-up quarter with no deal above $22M and a focus on early enterprise AI security platforms. |
| Q2 2025 | 6 | $96.3M | Aurascape's $50M seed round drove most of the quarter's total, masking modest activity elsewhere. |
| Q3 2025 | 5 | $230.6M | Noma Security and Airia each raised $100M, making this the first quarter with mega-deal dynamics. |
| Q4 2025 | 11 | $137.5M | Highest deal count of the period, but with smaller average checks as early-stage startups flooded in. |
| Q1 2026 | 5 | $273.25M | LMArena's $150M Series A drove a new record, with agentic AI governance themes dominating the narrative. |
| All quarters | 33 | $817.2M | Total for the 5-quarter period, with Q3 2025 and Q1 2026 together accounting for over 60% of all capital raised. |
In our AI safety market deck, we identify repeatable patterns you can use if you’re building in this market
Which startups in the AI safety market raised the largest rounds over the last months?
These startups raised the most recently in the AI safety market:
- LMArena raised $150M because investors bet that trusted, large-scale AI evaluation infrastructure will become mandatory as model proliferation makes safety comparisons essential.
- Noma Security raised $100M as enterprise demand for end-to-end AI agent security surged, covering everything from discovery to runtime protection in a single platform.
- Airia raised $100M from its own co-founder to accelerate its governance, security, and orchestration layer for enterprises scaling AI safely.
- WitnessAI raised $58M to expand globally and deepen its AI agent governance capabilities as agentic workflow adoption accelerated across enterprises.
- Aurascape raised $50M at seed stage to bring its AI Activity Control platform to market, targeting the growing problem of unmanaged shadow AI inside large organizations.
- Keycard raised $38M across seed and Series A to build identity and access infrastructure purpose-built for AI agents, which need task-scoped permissions rather than broad standing access.
- JetStream Security raised $34M to become the governance layer enterprises need as AI agents transition from pilot projects to full production deployments.
- Virtue AI raised $30M to bridge what it calls the critical AI security gap, offering a unified platform that spans red teaming, guardrails, and safety workflows.
- Fiddler AI raised $30M in a Series C to position itself as the control plane for AI agents, providing ongoing monitoring and governance for compound and agentic AI systems.
- GetReal Security raised $17.5M to scale its deepfake detection and AI-generated impersonation mitigation platform as governments and enterprises faced rising synthetic media threats.
And, yes, we do cover most of them in our beautiful pitch about the AI safety market.
You may also want to check our ranking of the most funded startups in the AI safety market as well as our list of the most valued startups.
In our AI safety market deck, we answer all the common questions from investors and entrepreneurs
Is the AI safety market shifting toward smaller or bigger deals?
Across all 33 deals from Q1 2025 to Q1 2026, the average AI safety market deal size was $24.8M, but that figure is heavily pulled up by a small number of very large rounds.
The average deal size per quarter ranged from $13.25M in Q1 2025 to $54.65M in Q1 2026, with the jump driven almost entirely by the two $100M-plus rounds in Q3 2025 and LMArena's $150M in Q1 2026. The trend line is rising, but it reflects a widening gap between a few large platform bets and many smaller early-stage AI safety plays.
If you strip out the top two deals from each quarter, the average AI safety market round stays quite modest, hovering between $8M and $15M, suggesting that the core of the market is still solidly early-stage rather than scaling toward large growth rounds.
| Quarter | Number of deals | Average deal size | Deals below $2M | Deals above $50M |
|---|---|---|---|---|
| Q1 2025 | 6 | $13.25M | 0 | 0 |
| Q2 2025 | 6 | $16.05M | 2 | 0 |
| Q3 2025 | 5 | $46.12M | 1 | 2 |
| Q4 2025 | 11 | $12.50M | 1 | 0 |
| Q1 2026 | 5 | $54.65M | 1 | 2 |
| All quarters | 33 | $24.76M | 5 | 4 |
In our AI safety market deck, we help you understand how the market is structured
How concentrated was funding activity in the AI safety market?
Funding in the AI safety market was highly concentrated throughout the entire period: in four of the five quarters, the single largest deal captured between 26% and 55% of all capital raised that quarter. This means that in any given quarter, a single company was able to claim a disproportionate share of total investment.
That concentration is even more striking when you look at the top three deals per quarter, which consistently accounted for 53% to 95% of all AI safety market funding, confirming that the market is not yet wide enough to distribute capital evenly across many large companies.
| Quarter | Number of deals | % by Top 1 | % by Top 3 | % by Top 10 |
|---|---|---|---|---|
| Q1 2025 | 6 | 26.4% | 66.0% | 100.0% |
| Q2 2025 | 6 | 51.9% | 92.4% | 100.0% |
| Q3 2025 | 5 | 43.4% | 94.7% | 100.0% |
| Q4 2025 | 11 | 27.6% | 53.1% | 99.3% |
| Q1 2026 | 5 | 54.9% | 88.6% | 100.0% |
| All quarters | 33 | ~18.3% | ~43.4% | ~85.0% |
In our AI safety market deck, we have designed useful charts to give you full market clarity
Which categories in the AI safety market received the most funding?
The AI security platform category raised $374M across 10 deals, capturing roughly 46% of all AI safety market funding, because investors are betting on broad platforms that combine visibility, governance, posture management, and runtime agent controls, rather than narrow point tools.
AI evaluation and red teaming attracted $207M across 9 deals (about 25% of total), but most of that total is driven by LMArena's $150M outlier round. Without LMArena, this category would average just $6.4M per deal, which suggests red teaming tools are seen as essential but not yet commanding platform-scale valuations.
AI governance and AI security posture management raised $175M across 10 deals (roughly 21% of total), and the category is broadening from pure compliance tooling toward operational control products that include real-time enforcement, audit trails, and production orchestration.
| Category name | Number of deals | Total raised | Startups and amount |
|---|---|---|---|
| AI security platform | 10 | $374.0M | GetReal Security ($17.5M), Straiker ($21M), Aurascape ($50M), Pillar Security ($9M), Noma Security ($100M), Airia ($100M), Keycard ($38M), Polygraf AI ($9.5M), Runlayer ($11M), Lumia Security ($18M) |
| AI evaluation / red teaming | 9 | $207.2M | Dreadnode ($14M), SplxAI ($7M), Yrikka AI ($1.5M), Repello AI ($1.225M), Promptfoo ($18.4M), AIM Intelligence ($1.3M), Haize Labs ($12.5M), LMArena ($150M), SecStrike ($1.25M) |
| AI governance / AI security posture | 10 | $175.0M | Singulr AI ($10M), Trustible ($4.6M), Modulos ($10.9M), Darwin AI ($15M), Portal26 ($9M), AI Score ($1M), Ciphero ($2.5M), WitnessAI ($58M), Fiddler AI ($30M), JetStream Security ($34M) |
In our AI safety market deck, we cover the latest tech updates shaping the market
Who are the biggest investors in the AI safety market?
Lightspeed Venture Partners is the most active investor in the AI safety market with 5 deals (Straiker, Virtue AI, Truth Systems, LMArena, and Fiddler AI), reflecting a deliberate strategy to build exposure across the full stack from guardrails to evaluation to governance.
Andreessen Horowitz participated in 3 deals (Promptfoo, Keycard, and LMArena), with a clear bias toward AI infrastructure and agent security plays where the market opportunity is large and the technical moat matters.
Insight Partners backed 3 deals (Promptfoo, Darwin AI, and Fiddler AI), combining a bet on open-source AI security tooling with a public-sector AI governance play and a mature AI observability platform.
Alumni Ventures appeared in 2 deals (Trustible and Polygraf AI), showing interest in governance and defense-oriented AI safety applications.
Ballistic Ventures backed 2 deals (GetReal Security and Noma Security), focusing on AI-specific security platforms with enterprise and government relevance.
Evolution Equity Partners also invested in 2 deals (GetReal Security and Noma Security), making it a consistent co-investor alongside Ballistic in this part of the AI security landscape.
Felicis backed 2 deals (Runlayer and LMArena), signaling interest in both agent security infrastructure and AI evaluation platforms.
Gradient Ventures (Google's AI-focused fund) participated in 2 deals (Truth Systems and Vijil), focusing on AI guardrails and agent resilience.
In-Q-Tel backed 2 deals (Dreadnode and GetReal Security), both of which have clear national security and government relevance in the AI safety market.
Mayfield invested in 2 deals (Aurascape and Vijil), with exposure to both the shadow AI governance problem and the AI agent resilience challenge.
Disclaimer: this investor list may be incomplete; we focus on publicly disclosed lead and prominent recurring investors, so some frequent minority participants may be underrepresented. "Total funded" does not represent the amount personally invested by an individual investor. Instead, it refers to the aggregate amount raised across all fundraising rounds in which the investor participated.
| Investor | Number of deals | Total funded | Startups |
|---|---|---|---|
| Lightspeed Venture Partners | 5 | $102.0M | Straiker, Virtue AI, Truth Systems, LMArena, Fiddler AI |
| Andreessen Horowitz | 3 | $202.4M | Promptfoo, Keycard, LMArena |
| Insight Partners | 3 | $63.4M | Promptfoo, Darwin AI, Fiddler AI |
| Alumni Ventures | 2 | $14.1M | Trustible, Polygraf AI |
| Ballistic Ventures | 2 | $117.5M | GetReal Security, Noma Security |
| Evolution Equity Partners | 2 | $117.5M | GetReal Security, Noma Security |
| Felicis | 2 | $161.0M | Runlayer, LMArena |
| Gradient Ventures | 2 | $21.0M | Truth Systems, Vijil |
In our AI safety market deck, we track adoption trends and shifts in consumer behavior
Related blog posts
- All funding deals in the AI safety market
- Which startups have raised the most funding in the AI safety market?
