SVCI: "Why We Invested in Knostic" - Leading CISOs' Thesis on AI Security

Silicon Valley CISO Investments (SVCI) has released their comprehensive market thesis on AI security, titled "SVCI in Brief: Why SVCI Invested in Knostic." This in-depth analysis details why they identify AI access control as the most critical layer in securing enterprise AI adoption after over a year and a half of extensive market research by some of the industry's most experienced security leaders.

A Collective CISO Perspective on AI Security

Silicon Valley CISO Investments (SVCI) is a group of 60+ leading CISOs from Fortune 500 to cutting edge technology companies who collectively invest in and support promising cybersecurity startups. SVCI has examined the rapidly evolving landscape of AI security challenges. Their thesis provides rare insight into how enterprise security leaders are thinking about the unprecedented risks introduced by LLMs and generative AI technologies.

Read the brief on why SVCI invested in Knostic here.

The Gap: AI Adoption Outpaces Security Capabilities

A central finding in SVCI's analysis identifies a critical and widening gap in the market:

"LLMs are reshaping how we work more profoundly than any technology in modern history... the gap between AI adoption and our ability to secure it is wider than anything we've seen. The risks are unmitigated, and vendor solutions remain scarce. Traditional security frameworks no longer apply."

This perspective highlights why so many enterprises struggle with securely implementing tools like Microsoft Copilot, Google Gemini, Glean, and other LLM-powered applications. The report details how existing security approaches fail to address the unique challenges of AI systems.

Why Access Control Emerged as Critical

Through their research, SVCI identified several foundational capabilities that any effective AI security solution must provide:

1. Context-aware access control that replaces traditional binary permissions
2. Data classification that understands semantic relationships, not just static tags
3. Zero-trust principles that extend to every AI interaction and output
4. Continuous validation instead of periodic certification for AI systems
5. Governance embedded into AI workflows, not as an afterthought

Particularly interesting is SVCI's assessment that current approaches to data security — including legacy DLP and access control solutions — cannot address these requirements because they "can't understand context, content, or the nuanced permissions required in AI systems."

Organizations Facing Immediate Challenges

The report identifies specific types of organizations experiencing the most urgent need for AI security solutions:

• Large enterprises with extensive Microsoft 365 deployments
• Departments handling sensitive data (Legal, HR, Finance)
• Security operations teams responsible for data loss prevention
• CISOs and privacy officers managing regulatory compliance

According to SVCI, these stakeholders face a common challenge: enabling AI-driven productivity while preventing sensitive data exposure.

Market Evolution and Competition

SVCI's analysis also examines how the AI security market is evolving, noting:

"There is considerable consensus among the SVCI group that the LLM access control problem will be prolific, impacting almost every enterprise across the globe over the next ten years."

The thesis contrasts different approaches to the problem, noting limitations in both existing security vendors and platform providers:

• AI Security Vendors (Firewalls and DLP): Most focus on detecting inappropriate content but lack context about the requestor for intelligent content moderation
• LLM Platform Providers: Primarily focused on productivity rather than security, creating inherent conflicts in their approach to access restrictions

Why SVCI Chose Knostic

After evaluating numerous approaches and startups in the AI security space, SVCI concluded their extensive market research by selecting Knostic as their investment of choice.  Huge thank you to SVCI for all the support. 

The thesis document explains several key factors that distinguished Knostic from others in the field:

A Differentiated Approach to AI Security

SVCI highlights Knostic's unique solution that provides:

• Context-aware access control that understands business context related to data, end-users, and intent
• Top-down data sensitivity focus enabling quick remediation of sensitive topics exposed through tools like Copilot
• Flexible deployment models with both SaaS and self-hosted options
• Centrally managed policy decision point that captures need-to-know rules across multiple LLM applications and agents

Looking Forward

The SVCI thesis concludes with confidence in their selection of Knostic:

"The AI security market is growing rapidly, fueled by enterprise adoption of generative AI and rising concerns around data protection. As companies embed AI applications and agents into core workflows, access control has emerged as the most critical control for managing risk."

"With a seasoned team, strong industry relationships, and a vision suitable for the AI age, Knostic is well-equipped to shape the future of AI security and capture both early adoption and long-term mindshare."

This newly released thesis provides valuable insight for CISOs, security professionals, and technology leaders navigating the complex intersection of AI adoption and security requirements, while underscoring why these experienced security leaders chose Knostic as their strategic investment in this critical space.

About SVCI: Silicon Valley CISO Investments (SVCI) is a group of 60+ leading CISOs from Fortune 500 to cutting edge technology companies who collectively invest in and support promising cybersecurity startups.Learn more at svci.io

About Knostic: Knostic provides AI access control that ensures LLM-based enterprise applications can provide employees with everything they need to know, but personalized within their need-to-know boundaries according to enterprise policy. Learn more at knostic.ai