Confident Security, a San Francisco-based startup, has raised US$4.2 million in seed funding led by Decibel, South Park Commons, Ex Ante, and Swyx.

The company’s product, Confsec, is an encryption tool designed to protect user data when interacting with AI models.

It prevents the storage or use of prompts and metadata for AI training purposes by the model provider or third parties.

Confsec applies encryption similar to Apple’s Private Cloud Compute, anonymizing data and allowing decryption only under strict conditions.

Founder and CEO Jonathan Mortensen said Confsec is aimed at privacy-sensitive sectors like healthcare, finance, and government.

The company is currently in discussions with banks, browsers, and search engines for potential integration.

The tool also logs AI inference processes publicly to allow expert verification of its privacy protections.

🔗 Source: TechCrunch

🧠 Food for thought

1️⃣ The rising financial impact of data breaches is driving demand for AI privacy solutions

The cost of data breaches has reached an all-time high in 2024, increasing by 10% year-over-year according to IBM’s latest research, making privacy-focused solutions increasingly essential for businesses adopting AI1.

Organizations handling sensitive information face particularly steep costs, with breached data stored in public clouds incurring the highest average breach cost at $5.17 million1.

These financial risks are compounded by the fact that 40% of data breaches now involve “shadow data”—information that organizations may not even know they possess—making protection increasingly complex1.

For regulated industries like healthcare and finance that Confident Security is targeting, these financial risks directly impact AI adoption decisions as they weigh innovation benefits against potential breach costs.

Organizations that implemented AI and automation in their security prevention reported average cost savings of $2.22 million compared to those that didn’t, demonstrating the financial incentive to address data privacy and protection proactively1.

2️⃣ Healthcare faces uniquely complex privacy challenges with AI adoption