
Security is easy to claim and hard to build. Most software companies start with a product and add security as they grow, patching in controls when a customer asks or an auditor requires it. Forsure took the opposite approach with SureSystem®, its agentic AI platform for employer healthcare and benefits decisions. The security architecture was designed alongside the product from day one.
That choice matters more than it might seem. When a platform handles claims data, employee health information, and financial decisions, security that is retrofitted tends to leave gaps. Security that is designed in tends to hold.
Layer One: Data Protection
At the foundation, all data in SureSystem® is encrypted in transit and at rest using industry standard protocols. Each customer’s data is logically isolated across Forsure’s multitenant infrastructure, so one organization can never reach another’s information. And Forsure draws a firm line on ownership: customer data is used to deliver the service to that customer, and it is never used to train foundation models or sold.
Layer Two: Access Control
Access is governed by role based and attribute based permissions with least privilege defaults. Every request is authenticated, authorized, and logged. In practice, that means people and systems see only what they need, and there is a record of every time they looked.
Layer Three: Governed AI
SureSystem® runs two agents, Alex for employers and Ollie for employees, that observe, decide, and act without waiting to be asked. That autonomy is what makes the platform useful, and it is also why governance is built into the architecture rather than written into a policy document. Both agents operate within defined guardrails, with human accountability for every consequential decision, and Forsure’s AI governance program is aligned to the NIST AI Risk Management Framework.
Layer Four: Continuous Verification
The environment is monitored on an ongoing basis, with regular audits and independent assessments. Those controls map to recognized frameworks: SOC 2, covering the security, availability, and confidentiality of Forsure’s systems; HIPAA, with Business Associate Agreements available; the NIST AI RMF; and TX-RAMP Level II for Texas public sector clients. HiTrust certification is currently underway.
Layer Five: Transparency
Finally, Forsure treats verification as something customers should be able to see for themselves. Its Trust Center gives security, privacy, and procurement teams access to certifications, policies, the subprocessor list, and an architecture overview, with deeper documentation available under NDA for active reviews. Forsure also welcomes independent security reviews, questionnaires, and control walkthroughs from prospective customers. When a customer leaves, its data can be exported and is then deleted in line with Forsure’s retention policy and contractual commitments.
Why This Approach Matters
IBM’s 2026 Cost of a Data Breach Report found that healthcare remains the most expensive industry for a breach, and that supply chain compromises, the kind that come in through a trusted vendor, were among the hardest breaches to detect and contain. A platform that makes decisions on cost, risk, and care has to be at least as trustworthy as the decisions themselves. Forsure’s view is that security, privacy, and governance are core product, not paperwork. The full SureSystem security architecture is outlined on Forsure’s Security and Trust page.
The Bottom Line
The best security architecture is the one a customer never has to think about because it was built right from the start, and can verify whenever they want to. That is the standard SureSystem® was designed to meet.