Bain Capital Ventures Funding Backs Risk Tools for AI-Driven Voice, Video Threats
A startup led by the ex-CEO of marketing platform Attentive raised $81 million to tackle the growing threat of artificial intelligence-powered social engineering and deepfake attacks.
See Also: Going Beyond the Copilot Pilot – A CISO’s Perspective
Adaptive Security plans to use the Bain Capital Ventures-led Series B funding round to grow its product, design and engineering teams in New York City and enhance its risk assessment, attack simulation and training capabilities, said co-founder and CEO Brian Long. He said Adaptive Security distinguishes itself with AI-driven simulations across SMS, voice and video, enabling more realistic attack scenarios.
“We’ve seen a large increase in AI-powered social engineering, in particular, huge increases in deepfake attacks,” Long told Information Security Media Group. “There was a 17x increase from ’23 to ’24, and we continue to hear consistently from customers every day that they’re seeing a large increase in these sophisticated deepfake, AI-powered social engineering attacks.”
Adaptive Security, founded in 2023, employs 154 people and has raised $136 million, finishing a $55 million Series A funding round in April 2025 led by Andreessen Horowitz and the OpenAI Startup Fund. The company has been led since its inception by Long, who served as CEO of SMS and email marketing platform Attentive for seven-and-a-half years and spent 19 months leading the mobile teams at Twitter (see: How AI Deepfakes Turn Human Trust Into a Cyber Risk).
From Email Phishing to SMS, Voice and Video Deepfake Threats
While traditional phishing via email has long been a focal point for cybersecurity teams, Adaptive Security sees a shift in the threat landscape toward more personalized, AI-enabled attacks across multiple channels including SMS, voice and video. Long also expressed concerns about data exfiltration, as employees increasingly engage with large language models that may unknowingly absorb and leak sensitive company information.
“I think you want to demonstrate two things: the monetary value you can bring to your team, and then number two, you want to be able to lead with a big vision,” Long said.
Today’s market for top-tier tech talent is highly competitive, Long said, with candidates getting pulled toward large tech firms, AI foundation model companies, or new startups. To stay competitive, Long said Adaptive must not only offer financial incentives but also project a bold, long-term vision that attracts the best minds in the field.
“We just wanted to make sure that we had the capital to invest in the top talent,” Long said. “The talent atmosphere right now is very competitive, between people going to the foundational models for AI or the prior generation technology companies or even new startups. It’s a very competitive environment.”
Adaptive Security currently evaluates more than 1,000 open-source intelligence data points per employee, surfacing information from public websites, social media activity and fitness tracking services that attackers could use in an AI-powered social engineering campaign. Adaptive wants to expand from 1,000 to more than 10,000 signals, massively increasing the system’s ability to detect obscure or long-tail risks.
“You might find only 10 or 20 really concerning things,” Long said. “But you want to make sure that you’re taking action to get rid of those things.”
What Makes Adaptive’s Security Awareness Training Different
SMS-based phishing tends to have significantly higher engagement because people read and respond to texts far more frequently than they do emails. Adaptive supports real-time voice deepfake simulations, allowing companies to simulate a call from any internal team member with just three seconds of audio. Using just a couple of minutes of footage, Adaptive can create a video avatar of any team member.
“A company can load up any person at their company with as little as three seconds of an audio sample to do a deepfake voice simulation,” Long said. “And it’s not scripted. It can be us calling them, us leaving a voicemail, or them calling us. I was just having a conversation with myself in Spanish yesterday, and it turns out that the avatar version of me is much better at Spanish than I am.”
While many companies offer security awareness content, Adaptive aims to make training personalized, AI-generated and context-aware, Long said. Based on user input and policy documents, Long said Adaptive can generate full training modules – including narration, video, visuals, quizzes and voiceovers – within minutes that are tailored to the employee’s role, recent failures, or even specific departments.
“We also have this crazy AI content creator that allows you to create a new training on anything in just a couple minutes,” Long said. “Say you want to do a training on advanced voice attacks for healthcare companies targeted to nurses. We can use that in three minutes to make a full new training with all the images and animations and videos, and even a quiz and full audio iteration just off that problem.”
Long said KnowBe4 dominates the security awareness training market with over an 80% share, followed by Proofpoint as a distant second, but he sees a clear opportunity for Adaptive to disrupt legacy platforms built for traditional phishing and static compliance content. Adaptive’s platform is focused entirely on AI-era threat vectors such as voice deepfakes, SMS-based scams and real-time avatar impersonations.
“Number one, we’re focused on the growth of all these different AI threats,” Long said. “Number two, we’re building it as modern software. And then number three, it’s a different age of building software with all these new, amazing tools. We can just build software that is much more flexible to a particular organization’s needs.”