In a recent episode of The Building Texas Show, hosted by Justin McKenzie, RocketDocs CEO Perry Robinson addressed a critical issue in enterprise AI: 80% of boards push AI adoption, but only 20% of companies trust the tools enough to deploy them. The episode, titled "Why 80% of Companies Don't Trust AI (And They're Right)," explores the widening gap between board mandates and operational reality, particularly in regulated industries like life sciences, healthcare, insurance, and financial services.
Robinson emphasized the risks of 'shadow AI,' where employees use free versions of ChatGPT, Claude, or Gemini with proprietary data, exposing companies to data leakage. He warned, "if you're not paying for the product, you are the product," noting that free tiers often train vendor models. This concern is amplified by policy shifts from major SaaS vendors, such as Atlassian's recent changes to training on customer Jira and Confluence data.
The conversation also covered the impending EU AI Act, which introduces revenue-based fines for non-compliance, and Salesforce's headless data moves reflecting pressure to feed AI pipelines. Robinson underscored that "policy is a promise, architecture is a guarantee," arguing contractual language alone cannot protect IP once employees route sensitive data through public models.
RocketDocs' Luma platform offers a secure generative AI layer running entirely on customer knowledge bases within their VPC, audited against ISO 27001 and SOC 2 Type 2 standards. Robinson described Luma as "limited on purpose," refusing to crawl the open internet to keep answers grounded in approved content. The platform also includes secure file transfer for defense, law enforcement, and product launch scenarios. Buyers increasingly include AI governance committees, chief compliance officers, and general counsel negotiating AI addenda.
