Security & Compliance of Clive™

Clive keeps claims compliant, encrypted, and audit-ready.

The idea of using AI in claims insurance is undeniably exciting. Having a claims-savvy adjuster always available, eliminating the need to sift through endless paperwork or hunt for minute details is a transformative vision, but it also raises an important issue – trust.

Claims data is highly confidential, containing sensitive policyholder information. Moreover, claims must be handled with meticulous precision, ensuring decisions are based on accurate, reliable data. Trusting a machine to perform such critical tasks doesn’t come naturally.

That’s why we designed Clive to be a claims adjuster you can trust. With advanced technology and strong safeguards, Clive is equipped with comprehensive guardrails, ensuring secure, precise, and compliant claims handling. You can rest assured—your claims management is in the best hands, even when they’re digital.

Clive is built on top of Google’s Enterprise-Grade Cloud Platform

Clive’s architecture makes him a secure, trusted adjuster for modern claims Management

Five Sigma maintains a strong, long-standing partnership with Google Cloud Platform (GCP). This partnership enables Five Sigma’s cloud-based products (Claims Management Platform and Clive) to scale securely in enterprise environments while fully adhering to the stringent regulatory, privacy, and security requirements of our customers. All customer data is safeguarded within GCP, ensuring comprehensive encryption both at rest and in transit.

Clive leverages Google’s Vertex AI to securely operate Large Language Models (LLMs) at scale, ensuring all processes are secure and reliable. This means Clive can provide adjusters with advanced AI-driven insights and recommendations while maintaining strict privacy standards—customer data is never used to train GCP’s models

By utilizing Google’s Gemini LLM models, Clive ensures that inputs and outputs remain confidential, with no logging or tracking. 

Additionally, GCP’s regional boundaries guarantee that all data processed by Clive stays within the designated region, eliminating cross-regional transfers and ensuring compliance with data residency regulations. 

Read Google's Case Study

Google has released an in-depth case study
about Clive™, Five Sigma’s revolutionary
AI Insurance Claims Adjuster

 

 

Model and Chat Compliance

Leveraging state-of-the-art LLM models with GCP’s configuration, Clive can screen both incoming questions, as well as the results they produce. Clive is capable of screening for:
  1. Hate speech
  2. Dangerous content
  3. Harassment
  4. Sexuality content
  5. Non-claims related content
Additionally, GCP applies its own screening process to questions and prompts. Outputs generated by Clive are filtered to ensure they do not include financial recommendations or the unwarranted content above

Watch Google’s movie about Five Sigma’s AI technology

Data Segregation & Claims Compliance

The data for the models is segregated between accounts and sub-accounts, so no data passes between different customer accounts. Data, queries, and answers will never be used outside of your organization’s account. Clive adheres to strict industry-leading standards, including SOC 2 Type 2 and GDPR

More To Explore

Browse our blog posts for the latest insights and trends in claims management.

Discover how to maximize the ROI on GenAI claims management tools by harnessing their ability to resolve claims faster and more efficiently. In his latest blog, our Senior AI Product Manager, Avi Yaacov, shares measurable ROI insights from Clive™, Five...
Read our latest blog to discover how Clive™, Five Sigma’s AI Claims Adjuster, empowers Chief Claims Officers with enhanced control, visibility, and decision-making tools. Learn how AI can boost efficiency without compromising oversight....
Learn about the risks of using public AI models like ChatGPT in insurance claims management and why they pose serious data privacy threats. Discover how you can harness the benefits of AI securely with industry-specific, compliant models designed to protect...