Skip to content

Adaptive Security

AI-driven platform that simulates deepfake, voice, and multichannel social-engineering attacks to train and test organizations against next-generation phishing.

Visit Website ↗ + Add to Compare
78/100Meaningful Innovator

Overview

Adaptive Security builds an AI-powered security awareness and simulation platform focused specifically on generative-AI social engineering: deepfake video and voice, AI-written phishing email, vishing, and smishing. Rather than relying on the static, template-based phishing simulations used by legacy security-awareness vendors, Adaptive generates realistic, multichannel attack scenarios modeled on how actual AI-enabled threat actors operate, then trains employees against them.

Founded in 2024 by Brian Long and Andrew Jones and headquartered in New York City, the company launched publicly in January 2025 and has since raised more than $145 million across three rounds, including a $55 million Series A anchored by the OpenAI Startup Fund and an $81 million Series B led by Bain Capital Ventures with participation from NVIDIA’s NVentures, Andreessen Horowitz, Capital One Ventures, and Citi Ventures. It counts more than 500 named enterprise customers, including PayPal, Figma, Xerox, the NHL, and the PGA of America.

The differentiator is narrowness and specificity: rather than a general awareness-training suite, Adaptive is purpose-built around the specific new attack surface generative AI has opened up — synthetic voices impersonating executives, AI-generated deepfake video, and personalized phishing at machine scale — a threat category that is expanding faster than most legacy awareness programs were designed to address.

Innovation Matrix Assessment

Innovation Velocity 8/10

Three funding rounds and rapid platform expansion from a single deepfake-simulation feature to a multichannel (email, SMS, voice, video) training platform within about 18 months of public launch.

Operational Value 8/10

Directly addresses a real, growing operational gap — employees are unprepared for synthetic-voice and deepfake-video impersonation, which legacy phishing-simulation tools don't cover. The named enterprise customer base (PayPal, Figma, Xerox) suggests the operational fit is translating into real adoption, not just a compelling pitch.

Market Momentum 8/10

Independently reported: $145M+ raised from OpenAI Startup Fund, NVIDIA, a16z, Bain Capital Ventures, Capital One Ventures and Citi Ventures; 500+ named enterprise customers per SiliconANGLE, including PayPal, Figma and Xerox.

Category Disruption 7/10

Security-awareness training is a mature category, but purpose-built deepfake/voice-clone simulation is a meaningfully different product than static phishing templates. The purpose-built, multichannel (voice/video/SMS) simulation approach is a genuinely different training model from static phishing simulators, warranting a modest upward revision.

Real-World Efficacy 8/10

The scale and profile of named enterprise customers (PayPal, Figma, NHL, PGA), independently corroborated by press coverage, is stronger real-world adoption evidence than most seed/early-stage security tools can show, even though formal independent efficacy testing hasn't been published yet; revised upward to reflect the strength of that adoption signal.

Enduring Relevance 8/10

AI-generated voice and video impersonation is a fast-growing, well-documented attack vector (executive fraud via cloned voices) that is likely to keep expanding as generative AI tools improve, keeping this category relevant for years.

Why CISOs Should Care

Gives security teams a way to test and train staff against deepfake and AI-voice impersonation specifically, a threat vector most existing awareness programs don't simulate at all.

What Makes It Different

Built natively around generative-AI attack techniques (synthetic voice, deepfake video, AI-personalized phishing) rather than adapting legacy static-template phishing simulators.

The Matrix Verdict

78/100 — MEANINGFUL INNOVATOR

A Meaningful Innovator still in its early scaling phase: strong, independently verifiable funding and named-customer momentum, and a genuinely different product angle on an old category, but real-world efficacy evidence is still mostly vendor-reported.

Editorial Note: Claims vs. Verified Findings

Customer-reported net promoter scores and specific efficacy percentages published on Adaptive's own case-study pages are vendor claims; the funding amounts, investor names, and named customers are corroborated by independent press coverage (SiliconANGLE, CNBC, PR Newswire).

Sources