Abnormal Security's $250M Series D made behavioral-AI email defense the loudest category bet, and the AI-native SOC race has since pulled in Prophet Security, Dropzone AI, and Reach Security as Tier-1 alert-triage agents. Cyera and Skyflow attack the data-security-posture problem from opposite angles, classification versus runtime tokenization, while Wiz, Snyk, and Vectra AI push existing CNAPP, AppSec, and NDR product lines toward agent-driven detection. HiddenLayer and Lakera carve out model-and-prompt security specifically. Adaptive Security flips defense outward by training employees against AI-generated phishing, a workload that didn't exist before mid-2024 deepfake voice scams.
CATEGORY
Best for Cyber Defense AI Tools
99 tools compared · 2026
AI-native SOC agents, model-supply-chain scanners, and behavioral defenses against deepfake-era attacks
99 ai for cyber defense startups tracked, with the largest concentration in US. Total tracked funding: $9.4B.
Tracked
99
Total Raised
$9.4B
Countries
10
Active Deals
0
1 
Upwind IL $430M 2 
XBOW US $237M 3 
Noma Security IL $132M 4 
Snyk Boston, GB $1.3B 5 
WitnessAI Atherton, US $85.5M 6 
Torq Portland, US $332M 7 
Vega Security IL $185M 8 
Saviynt El Segundo, US $870M 9 
Oasis Security Tel Aviv, IL $195M 10 
Endor Labs Palo Alto, US $188M 11 
GitGuardian Paris, FR $106M 12 
Reality Defender New York, US $48M
Funding by year — AI for Cyber Defense
2019 → 2026Market overview
Key trends 2026
- AI-native SOCs replace Tier-1 analysts. Prophet, Dropzone, and Adaptive ship agents that close alerts without human triage.
- Prompt-injection gets its own product class. Lakera and HiddenLayer sell guardrails and red-teaming for the model layer.
- Identity-fraud meets deepfakes. Persona, Alloy, and Sardine retool KYC stacks to catch synthetic-media account-opening attacks.
Benchmarks vs global
Largest single round
$250M (Abnormal Series D)
vs traditional EDR rounds ↑
AI-SOC alerts auto-closed
60-80% (vendor-claimed)
vs ~10% pre-agent baseline ↑
Deepfake-phishing incidents YoY
3-5x growth
vs flat traditional phishing ↑
Top countries
By startup countStage breakdown
Latest round typeTop investors backing AI for Cyber Defense
See all →Y Combinator
15 deals
BV
Bessemer Venture Partners
10 deals
Accel
8 deals
T
Team8
7 deals
Menlo Ventures
7 deals
Insight Partners
7 deals
Khosla Ventures
7 deals
EE
Evolution Equity Partners
5 deals
FC
Foundation Capital
5 deals
PC
Picture Capital
5 deals
Lightspeed Venture Partners
4 deals
A
Addition
4 deals
FAQ
Frequently asked
What does AI-native SOC actually mean?
It means triage agents (Prophet Security, Dropzone AI) auto-investigate alerts end-to-end — pulling logs, correlating identities, drafting verdicts — instead of routing tickets to a human Tier-1 analyst. The SOC keeps Tier-2 and Tier-3 humans for escalations only.
Which AI cyber-defense startup has raised the most?
Abnormal Security leads our tracked set at $250M Series D, focused on behavioral AI for email and account takeover. Adjacent CNAPP leaders like Wiz are larger overall but sit primarily in cloud security rather than AI-defense-first.
Do I need a separate tool for AI/LLM security on top of CNAPP?
Most teams now run both. CNAPP (Wiz, Snyk) covers cloud and code. HiddenLayer and Lakera add model-supply-chain scanning, prompt-injection detection, and runtime guardrails for deployed LLM apps that traditional CNAPP tools don't yet inspect.
Recent rounds in AI for Cyber Defense
All rounds →| Date | Startup | Round | Amount |
|---|---|---|---|
| May 2026 | Exaforce | Series B | $125M |
| May 2026 | Frame Security | Seed | $50M |
| May 2026 | XBOW | Series C Extension | $35M |
| Apr 2026 | Artemis Security | Series A | $55M |
| Apr 2026 | Trent AI | Seed | $13M |
| Mar 2026 | Oasis Security | Series B | $120M |
| Mar 2026 | XBOW | Series C | $120M |
| Mar 2026 | RunSybil | Series A | $40M |
All AI for Cyber Defense startups
Page 4Riverbank
AI-native Red Teaming and Offensive Security
Raised
$500K
Stage
Seed
Ocean Security
AI security agents that investigate every email's intent in real time
Raised
$28M
Stage
Seed
Knostic
Need-to-know access controls and security for enterprise AI
Raised
$14M
Stage
S-A
imper.ai
Real-time defense against AI-driven impersonation and deepfake attacks
Raised
$28M
Stage
Seed
Acoru
Predict and prevent AI-driven fraud and money mules before they strike
Raised
$10.8M
Stage
S-A
TruthScan
Enterprise AI detection for deepfakes, text, images, and voice
Raised
$16.2M
Stage
S-A
Kindo
The only On-Prem, AI-Native Agent Harness for Enterprises, designed for technical operations.
SandboxAQ
Transforming the world with AI and advanced computing, applying quantitative AI to solve complex challenges in various industries.
Cyera
AI-Native Data Security for Cloud, SaaS, On-Prem, and AI environments.
Reach Security
Proactively identify and remediate security control misconfigurations to close gaps faster than AI-powered attacks.
Prophet Security
Empower your SOC with a fleet of autonomous AI agents that accelerates Tier 1, Tier 2, and Tier 3 tasks.
Dropzone AI
AI Agents for Alert Triage & Threat Hunting, enabling your SOC to operate at machine scale.
Adaptive Security
Security Awareness Training Built for AI Threats.
Sardine
The top agentic risk platform for real-time fraud prevention, AML compliance, and AI-driven attack protection.
HiddenLayer
Total AI Security for the enterprise.
Wiz
AI Cybersecurity for All Your Cloud and AI Applications.
Coram AI
AI-native physical security platform for video, access and safety
Raised
$13.8M
Stage
S-A
AegisAI
Autonomous AI agents that stop phishing before it reaches the inbox
Raised
$13M
Stage
Seed
CodeAnt AI
Agentic AI code review and application security across 30+ languages
Raised
$2M
Stage
Seed
Mate
AI-native SOC that learns your environment and runs investigations
Raised
$15.5M
Stage
Seed
Salv
The fincrime platform with intelligence sharing built-in to detect money laundering, exchange intelligence, and stop fraud more effectively.
Lakera
The leading AI-native security platform to secure your AI future and accelerate GenAI, agents, and MCPs for enterprise teams.
CodeIntegrity
Runtime security guardrails for agentic AI
Raised
$5M
Stage
Seed
Unosecur
AI-driven identity security for human and non-human identities
Raised
$5M
Stage
Seed