Pillar, an innovative new platform, is designed to meet the increasing challenges of AI security. Cyber security experts Dor Sarig and Ziv Karlinger started Pillar to protect good users from bad bots. Their intention is to improve security at all stages of the AI lifecycle. By employing innovative strategies that extend beyond traditional frameworks, Pillar empowers organizations to effectively manage AI-specific risks and build trust in their AI systems.
The platform automatically gathers telemetry data from all aspects of the AI stack. This provides organizations with critical transparency to protect their AI deployments. To strengthen your security pillar leverages the industry’s largest AI-based threat intelligence feed. Combined with this, it empowers its data with more than 10 million real-world data points. This incredible abundance of data inspires it to find weaknesses and dangers with near-psychic precision. By being proactive, organizations can be one step ahead of cyber attackers looking for their next breach.
A Comprehensive Approach to AI Security
It’s this critical understanding of AI security that distinguishes Pillar from traditional approaches. What makes Pillar unique is its commitment to make security a priority at the outset of AI development. It commits to this focus continuously through deployment and operation, beyond the typical Software Development Life Cycle (SDLC) frameworks or CI/CD pipelines. In so doing, it shores up potential deficiencies that could occur in each stage of the AI lifecycle.
The platform uses established security framework models such as STRIDE, ISO, MITRE ATLAS, and the OWASP Top 10 for LLMs. This makes sure your data is well protected. Pillar has created its own SAIL framework, specifically tailored to address the distinct characteristics of AI systems. This framework builds on the best-known standards in the field, the GRI. Such a thoroughgoing methodology guarantees that organizations will be able to fully implement the necessary steps to identify and mitigate risks that AI technologies pose.
Additionally, Pillar has developed black-box, target-based red teaming services for enterprises adopting third-party and embedded AI applications. By arming themselves with this capability, organizations can simulate real-world attacks against their data and probe their data boundaries. By revealing exposure risks via these simulations, organizations can take proactive measures to harden their defenses against potential intrusions.
Enhancing Security During Operation
One big innovation that Pillar brought was a heavier emphasis on security, especially in the Operate phase of AI deployment. Pillar combines secure sandboxing techniques to maintain a frictionless experience for AI applications. This multistage, risk-based approach prevents critical infrastructure and sensitive data from being compromised. Focus agents on the mission. This approach is in harmony with the zero-trust principles that Pillar champions, isolating agents from critical systems to reduce risk exposure.
Pillar’s isolated runtimes improve security by letting agents work efficiently without compromising strict security protocols. This isolation passively prevents drawbacks from third-party applications. Very importantly then, it serves as a shock absorber protecting the larger organizational ecosystem from any harmful effects.
The tools for assessing and securing third-party AI systems have matured. As a big leap forward in security protocols, this new industry standard requires no integration or customization. Regain trust and accountability Organizations can confidently deploy AI technologies, knowing that Pillar’s robust framework protects their operations from emerging threats.
Leadership with a Wealth of Experience
Pillar’s leadership team combines extensive cybersecurity experience in both offensive and defensive techniques. In addition, co-founder and CEO Dor Sarig comes from a long pedigree in cyber-offensive operations. This practical experience provides him with intimate knowledge of the vulnerabilities his opponents will seek to exploit. His vision for Pillar focuses on giving organizations the tools to take ownership of their AI security landscape.
Ziv Karlinger, co-founder and CTO Karlinger rounds out the trio with more than a decade developing offensive and defensive techniques in the realm of cyberspace. His technical acumen allows Pillar to see threats before they happen. He provides tactical guidance that enables organizations to better fortify and protect themselves.
As a team, Sarig and Karlinger are at the helm of a team focused on empowering innovative organizations to tackle AI-specific risks responsibly. Their combined experience positions Pillar as a trusted partner for businesses looking to navigate the complex world of AI security.
