Cranium launches exposure management solution to safeguard AI systems – Business
Artificial intelligence security and trust software firm Cranium.ai Corp. today announced the launch of Cranium AI Exposure Management, designed to help protect and secure internal and third-party AI solutions.
Cranium AI Exposure Management features an AI-augmented workflow with a secure large language model architecture that has been paired with threat intelligence to provide visibility into a given AI system. The service characterizes attack surfaces and assesses vulnerabilities in an organization.
Identifying vulnerabilities is the key to the new offering, ensuring the security and reliability of machine learning applications. The service also empowers red-teaming efforts to discover novel threats, inform protection strategies and harden AI systems against known adversarial tactics and vulnerabilities to help secure AIs and LLMs in the development stages and in usage.
With Cranium AI Exposure Management, users can red-team a full OODA loop — the Observe, Orient, Decide and Act decision-making model developed by U.S. Air Force Colonel John Boyd. With the service, users can observe, map and characterize the AI attack surface, orient through the gaining of insights through threat modeling against AI objectives and policies, and decide and act with data and information on hand to take action on high-priority threats.
The solution further offers a comprehensive framework for the safe integration of AI capabilities into business operations, enhancing visibility across AI infrastructure to optimize utilization and mitigate risks while implementing strategies to address vulnerabilities. The solution adheres to current and future regulatory standards and holds vendors accountable for compliance with these evolving standards.
“With AI now embedded into all the systems and business processes that make our world work, organizations face a critical dilemma: No one knows if their AI is secure,” said Chief Executive Jonathan Dambrot. “Add in increasing regulatory pressure and the pace of AI advancement and risks are rapidly increasing. We’re solving these challenges by allowing organizations to gain peace of mind with an industry-first exposure management solution.”
Cranium, which launched as a standalone company from KPMG Studio in April 2023, was last in the news in October when it raised $25 million in new funding to enhance AI security and trust. Investors in the company include Telstra Ventures LLC, KPMG and SYN Ventures LLC.