Home
News
Tech Grid
Interviews
Anecdotes
Think Stack
Press Releases
Articles
  • AI

Confident Security Unveils CONFSEC for Private AI Inference


Confident Security Unveils CONFSEC for Private AI Inference
  • by: Source Logo
  • |
  • July 18, 2025

Confident Security, a pioneer in private AI infrastructure, has launched its enterprise-grade CONFSEC platform, designed to ensure provably private AI interactions. With $4.2 million in funding from Decibel, South Park Commons, Ex Ante, and Swyx, the San Francisco-based company addresses growing privacy concerns in AI adoption, particularly in sectors handling sensitive data like healthcare, finance, and government.

Quick Intel

  • Confident Security launches CONFSEC, a private AI inference engine.

  • Raised $4.2M from Decibel, South Park Commons, Ex Ante, and Swyx.

  • CONFSEC ensures prompts and metadata remain private and unencrypted.

  • Deployable on any cloud or bare-metal, enhancing flexibility.

  • Targets high-trust sectors like healthcare, finance, and government.

  • Led by Jonathan Mortensen, with expertise from Google and Apple.

CONFSEC: A Breakthrough in AI Privacy

Confident Security’s CONFSEC is an enterprise-grade implementation of Apple’s Private Cloud Compute (PCC) architecture, designed to protect sensitive data during AI inference. Thoroughly tested, externally audited, and production-ready, CONFSEC enables large language model (LLM) providers, hyperscalers, and enterprises to wrap AI inference engines, ensuring prompts and metadata are never used for training, viewed by third parties, or accessed in unencrypted form. “Businesses and consumers are feeding AI everything from medical information to a company’s roadmap and trade secrets,” said Jonathan Mortensen, Founder and CEO of Confident Security. “AI is now table stakes, but it’s come at the cost of privacy.”

Addressing Privacy Barriers in AI Adoption

As AI systems handle sensitive data like medical records and trade secrets, privacy has become a critical barrier, particularly in high-trust industries. CONFSEC uses advanced technologies such as OHTTP, blind signatures, remote attestation, and trusted execution environments (TEEs) to guarantee data privacy. This approach allows businesses to leverage AI’s competitive advantages without compromising intellectual property or customer data. “Privacy is now the critical barrier to AI adoption in enterprise,” said Jess Leao, Partner at Decibel. “Confident Security is ahead of the curve in recognizing that the future of AI depends on trust built into the infrastructure itself.”

Flexible Deployment for Enterprise Needs

CONFSEC’s versatility enables deployment on any cloud or bare-metal environment, offering businesses technical privacy guarantees beyond legal assurances. This flexibility makes it ideal for sectors like healthcare, finance, and government, where data security is paramount. By embedding privacy into AI infrastructure, Confident Security unlocks AI adoption in compliance-heavy environments, fostering trust and scalability.

Expert Leadership Driving Innovation

Led by Jonathan Mortensen, a two-time founder with exits to BlueVoyant and Databricks, Confident Security’s team includes experts from Google, Apple, Databricks, RedHat, HashiCorp, Stanford, and Johns Hopkins. Their deep expertise in secure systems and AI infrastructure positions Confident Security as a leader in private AI solutions.

Confident Security’s launch of CONFSEC marks a pivotal step in making AI trustworthy for enterprises. By prioritizing privacy through innovative technology, the company enables organizations to harness AI’s potential while safeguarding sensitive data, setting a new standard for secure AI deployment in critical industries.

 

About Confident Security

Confident Security is the creator behind CONFSEC, an enterprise-grade version of Apple’s Private Cloud Compute (PCC) - thoroughly tested, externally audited, secure, and production-ready, and deployable on any cloud or your own bare-metal. Using a combination of OHTTP, blind signatures, remote attestation, TEEs, TPMs, transparency logs, and more, Confident Security provably guarantees nobody can see the user’s prompt. The company is led by Jonathan Mortensen, a two time founder that sold companies to BlueVoyant and Databricks, and built by a team with deep expertise in secure systems, AI, infrastructure, and trusted computing backgrounds from Google, Apple, Databricks, RedHat, and HashiCorp.

News Disclaimer
  • Share