It’s 2 PM on a Friday when a patient calls your privacy hotline. She suspects someone accessed her medical records without reason. A year ago, this reactive investigation would have consumed a privacy analyst’s afternoon – manually pulling and cross-referencing access logs, without a clear lead. Now, the analyst opens Prism Assistant in PrivacyPro and types: “Can you find who accessed Jane Doe’s records outside of her care team in the last 90 days?” Moments later, the analyst has a complete access history, risk scores, and a plain-English summary of what to look at next. What used to take over an hour now takes just a couple minutes.
Bluesight Expands Prism Assistant to PrivacyPro
Prism, Bluesight’s AI platform for hospital pharmacy and compliance teams, was unveiled earlier in 2026 with its first product Prism Assistant available in ControlCheck. Prism Assistant has expanded to PrivacyPro, Bluesight’s patient privacy monitoring software, giving privacy analysts and investigators an AI assistant built specifically for patient privacy workflows. Investigators can ask questions about improper EMR accesses, risk patterns, and cases in natural language then receive evidence-backed answers, summaries, and charts in under a minute.
A Call for More Efficient Patient Privacy Operations
The math isn’t working. As data volumes, staff activity, and patient visits increase, PrivacyPro customers reviewed 46% more cases in 2025 compared to just two years prior. The average organization now handles 1,063 cases a year, more than 20 every week. But privacy teams haven’t grown at the same pace.
To continue protecting organizational risk for breached patient records, privacy teams need better tools. Bluesight research shows that a combined 69% of privacy staff report using AI for their work either on a daily or weekly cadence. However only 35% of healthcare organizations have visibility into how their staff use third-party AI tools. This gap shows that the demand for AI and the guardrails in place for AI may not match. If an analyst were to use a third party tool to analyze access data, they may be creating a new privacy problem while solving an old one.
Prism Assistant gives lean privacy teams a sanctioned place to ask those questions, analyze privacy investigation data, and build reports.
Health Systems Are Already Using Prism to Accelerate Investigations
Four health systems are already using Prism Assistant. Here’s what they’re using it for:
Reactive investigations. When a patient calls with a complaint or a vague hotline tip comes in, investigators ask Prism to pull every EMR user beyond their care team who accessed a patient in the last 90 days, sorted by suspicion score.
Understanding the “why.” One of the most popular questions: “Why was this access flagged?” Investigators and analysts get a breakdown covering the possible personal relationship, clinical context, and suggested next steps—which then shapes their follow-up questions.
Identity resolution. One privacy team couldn’t find an EMR user in the global email directory. Prism found them by name and role, with richer detail than the standard user tab, then checked for other suspicious accesses before reassigning the case.
Reporting upward. A compliance operations director built quarterly reports with Prism and shared them directly with the chief compliance officer. Another team saved a full Prism conversation and brought it straight into a meeting. A third health system raised its weekly case limits—faster review gave them room to handle more.
Early adopters report saving one to two hours of manual research per case. Across the 1,000+ cases health systems investigate through PrivacyPro annually, that translates to an estimated 20 to 40 hours saved every week – time teams can redirect toward higher-value investigations or increasing case capacity.
Increase Capacity without Jeopardizing Data Security
More capacity only helps if patient data stays protected. At the 2026 Privacy Symposium, Mustafa Rahimi of AWS presented what he called the trust equation, a way to evaluate any AI that works with sensitive data. It has four parts: identity, guardrails, transparency, and validation. He was clear that no single control is a silver bullet and that the four work as layers. They make a practical checklist for privacy officers and CISOs evaluating an AI tool that touches patient access data, so here is how Prism Assistant answers each one.
- Identity. Prism Assistant works through PrivacyPro’s existing security roles. Each user sees only the data they’re already authorized to see.
- Guardrails. Prism runs entirely inside Bluesight’s HIPAA-compliant AWS environment. No PHI leaves Bluesight’s private cloud and its single-tenant design keeps each health system’s data isolated. It doesn’t train on customer data and it answers only from your PrivacyPro data.
- Transparency. Every session is saved. Full conversations can be exported as PDF, CSV, or PNG, giving your program a record of how conclusions were reached.
- Validation. Privacy teams should take a “trust, but verify” approach to each answer, and can rate each response with a thumbs up or a thumbs down which gives Bluesight real-time feedback.
Learn More
To see Prism Assistant in PrivacyPro, visit bluesight.com/Prism.