Why we incubated & invested in WitnessAI

Revolutionizing safe AI adoption

By Barmak Meftah, Co-founder and General Partner

With a focus on early-stage cybersecurity startups, we’re in a unique position at Ballistic to not only invest in innovative companies but to incubate them from the ground up. As former operators, this gives my partners and I the continued opportunity to work on solving complex problems by identifying promising ideas and assembling the right team with the right expertise to bring those ideas to life. 

Today we’re excited to share that WitnessAI – our first publicly announced incubation – is launching from stealth with $27.5 million in Series A funding co-led by Ballistic and our peers at GV. WitnessAI solves a critical cybersecurity challenge that almost every company across every industry is thinking about right now.

Safe AI Adoption

The democratization of AI is here. We knew it was coming, but when ChatGPT acquired 1 million users just 5 days after launching in November 2022, we knew that along with democratization and its many promises, enterprises would face a new world of governance and security challenges. WitnessAI emerged from that vision, enabling enterprises to adopt AI both safely and effectively.

From its inception, we handpicked a talented team of experts from diverse backgrounds in AI, cyber and web-scale platform development to set WitnessAI apart. Led by CEO Rick Caccia, who has a history of bringing successful security products to market, the WitnessAI team believes that solving the secure-AI-usage problem requires three areas of expertise: AI model development, offensive and defensive security, and high-scale web services deployment. Without all three, no enterprise can deploy AI guardrails: Without the AI expertise, the system won’t get smarter as the LLMs it protects get better. Without the security expertise, your guardrails will be a weak point. And without the high-scale web platform, it will just be a bottleneck.

Securing LLMs vs. Uniquely Securing AI Usage

In what has quickly become a crowded market, WitnessAI stands apart by focusing on securing AI usage, rather than just securing the AI model itself. 

The WitnessAI Secure AI Enablement Platform provides comprehensive visibility, control and protection of enterprise AI usage. That includes modules for observability of AI use, policy enforcement and governance of AI use, and protection of employees, customers, and enterprise data from AI use. 

Only two other alternatives are out there today: legacy or other startup solutions. While legacy EDR, CASB and firewalls from big vendors may seem appealing to some, they fall short in understanding AI usage; instead, they block it – preventing speed and innovation that businesses need. On the other hand, almost all other startups trying to solve this problem focus on securing the LLM itself, not on securing AI use. Securing the use of the model is a real problem – and that’s why WitnessAI is the real solution.

We’re incredibly excited for WitnessAI and our shared vision to make AI safe and accessible and help businesses move fast and stay innovative. Be sure to check them out!