Securing AI: The Next Frontier with Action Firewalls

A Gujarat-based neuro-engineering deep-tech AI firm has unveiled an innovative ''AI Action Firewall'' intended to enhance the safety of artificial intelligence systems through pre-execution verification, according to an official statement issued Saturday.

This firewall serves as a security layer between AI systems and real-world operations, ensuring that actions are authorized, thoroughly monitored, and meticulously recorded. It expands on the concept of traditional firewalls but is uniquely tailored to regulate AI-driven activities such as sending emails, executing code, accessing databases, or initiating automated workflows, explained Saurabh Patel, Founder and CEO of NeuroPause Lab Limited, during a press briefing.

The firm has applied for a global patent for this technology, with each AI action subjected to predefined policies, categorized as 'allow', 'review', or 'block'. Routine tasks proceed automatically, while high-risk operations require human approval to ensure a permanent audit trail is created, contributing to organizational transparency and regulatory adherence, Patel added.