Unbound secures USD $4 million seed to boost safe AI adoption

Unbound has announced the close of an oversubscribed seed funding round, raising USD $4 million to support enterprise adoption of generative AI tools while maintaining security and governance controls.

The funding round was led by Race Capital, with participation from Wayfinder Ventures, Y Combinator, Massive Tech Ventures and several notable angel investors.

Other investors included Alpha Square Group, Northside Ventures, Liquid2, Pioneer Fund, Scale Asia Ventures, SBXI, alongside individual investors such as Ram Shriram, Dr. Trishan Panch, Dr. John Brownstein, Taro Fukuyama, Eli Brown, Chris Siakos, Joe Vadakkan, Zain Rizavi, and Finbarr Taylor.

Unbound provides IT teams with a platform that integrates into commonly used generative AI tools, offering real-time protection, model routing, and usage analytics. The AI Gateway product aims to prevent sensitive data from being leaked, manage the costs associated with AI models, and enable controlled rollouts of new AI models within large organisations.

The company's CEO and Co-Founder, Rajaram Srinivasan, brings experience from previous roles overseeing data security products at Palo Alto Networks and Imperva.

CTO and Co-Founder Vignesh Subbiah was formerly a founding team member at Tophatter and Shogun and previously scaled engineering teams through periods of significant growth. The pair co-founded Unbound after identifying security risks arising from rapid adoption of AI models such as GPT-3.5, where teams were found to be sharing sensitive information with little to no oversight.

Unbound's approach differs from existing data loss prevention tools by enabling real-time redaction of sensitive data and automatically routing high-risk prompts to internally hosted, open-source models. According to the company, this has already resulted in the prevention of hundreds of secret credentials leaks and more than 500 incidents involving personal information, including customer names, phone numbers, and patient records.

The company also offers cost control features, permitting organisations to allocate premium AI models to high-stakes workflows – such as engineering teams developing core infrastructure – and route less critical tasks to lower-cost open-source solutions. Unbound reports that mid-market customers have achieved annual savings of more than USD $10,000 on AI seat licenses since adopting the platform.

Existing customers span sectors including technology and healthcare. One technology company recently used Unbound to facilitate the safe introduction of Gemini 2.5 AI tools into production environments for over 100 engineers in a single week.

Srinivasan commented, "As AI tools become mainstream, enterprises are turning to flexibility and control. They want visibility into what's being used, assurance that their data is protected, and the ability to swap in better models as the space evolves. Unbound is the bridge that makes that possible."

Subbiah added, "Defaulting to blanket bans on AI tools is like being in the times of GPT 3.5. Unbound enables surgical security controls into every AI request so teams can innovate freely without putting corporate secrets at risk. In just a few months, our customers have prevented over 7,000 potential data leaks and cut AI tooling costs by nearly 70 percent."

Abraham Ingersoll, Chief Information Security Officer at The Hut Group, described the company's adoption of Unbound: "At THG Ingenuity, we see the security team as an enabler, not a blocker. Unbound empowers us to roll out AI tools to employees with confidence. Unbound AI Gateway's data protection controls and intelligent routing have been instrumental in safeguarding sensitive data while helping us optimize costs."

Edith Yeung, General Partner at Race Capital, noted the market opportunity: "AI is projected to reach USD $4.8 trillion in market value for the enterprise by 2033 globally — but without proper guardrails, that value is at risk. From shadow models to data leaks, the dangers of unmanaged AI are very real. We are excited to back Rajaram Vignesh and the Unbound Security team as they create a new category of AI infrastructure: one built for safety, observability and cost discipline from day one. We're proud to back Rajaram, Vignesh, and the team building a new category of AI infrastructure – one that makes enterprise adoption safe, observable, and cost-efficient from day one."

Unbound plans to use the fresh capital to expand its integrations across the AI ecosystem, deepen software routing capabilities, and support internal model orchestration for enterprises pursuing open-source large language models.