Is Gemini HIPAA Compliant? Workspace, Vertex AI, and the BAA in 2026
Sonomos Research · Sonomos
The Sonomos research team writes about AI privacy, data protection, and how to use generative AI safely at work.
Short answer: Gemini is not "HIPAA compliant" out of the box. Google does enter into Business Associate Agreements (BAAs), and Google's HIPAA guide says its Google Cloud BAA covers listed Google Cloud services and listed Google Workspace services, including the Gemini app (excluding Gemini in Google Chrome), Gemini in Google Workspace, and the Gemini Mac app. Consumer Google accounts have no BAA with Google, so protected health information (PHI) does not belong in them. The guide also puts the responsibility on you: you decide whether you need a BAA, and a service you cannot confirm is supported must not be used for PHI. This guide explains, in 2026, what HIPAA actually requires of an AI tool, which Google products can be made compliant, and how to keep PHI from leaking when clinicians and staff use unsanctioned accounts.
Last reviewed October 1, 2026 against the regulations and vendor documentation listed under Sources and primary references.
What HIPAA requires of any AI tool
HIPAA is implemented through three rules that all attach when an AI vendor processes PHI on your behalf:
- Privacy Rule — defines PHI and the conditions under which it may be used or disclosed.
- Security Rule — defines administrative, physical, and technical safeguards for electronic PHI (ePHI).
- Breach Notification Rule — defines what counts as a breach and the timelines for reporting it.
For Google to lawfully process PHI for a covered entity using Gemini, three things must be true:
- Google qualifies as a business associate because it creates, receives, maintains, or transmits PHI on the covered entity's behalf for a regulated function or activity.
- The covered entity has a signed Business Associate Agreement (BAA) that contractually extends HIPAA obligations to Google and meets the content requirements of 45 CFR 164.504(e).
- The Gemini product, in the configuration you actually use, supports the Security Rule's technical safeguards: access control, audit controls, integrity, authentication, and transmission security, with encryption an addressable specification.
A covered entity may disclose PHI to a business associate only after obtaining satisfactory assurance that it will safeguard the information, documented in a written contract (45 CFR 164.502(e)). Without that, sending PHI to Gemini is an impermissible disclosure, no matter how robust Google's underlying infrastructure is.
Where Google stands on HIPAA in 2026
As of October 1, 2026, Google's HIPAA compliance guide, a page Google updated on September 30, 2026, describes a single Google Cloud BAA that covers both Google Cloud and listed Google Workspace services. Earlier versions of this guide split Workspace and Cloud into separate BAAs; Google's current documentation does not.
- Gemini in Google Workspace and the Gemini app. The guide's list of Workspace services included in the Google Cloud BAA names the Gemini app (excluding Gemini in Google Chrome), Gemini in Google Workspace, and the Gemini Mac app, alongside Gmail, Drive, Meet, Chat, and others. Google dates that list August 31, 2025.
- Gemini on Google Cloud. The Cloud services list names Gemini Enterprise, Gemini Enterprise Agent Platform, Generative AI on Agent Platform, Gemini Code Assist, Gemini in BigQuery, Gemini in Colab Enterprise, and Gemini Notebook Enterprise. Google's product names change, so match the service you intend to use to the current list by name. Earlier versions of this guide referred to Gemini on Vertex AI.
- What is outside the BAA. The guide says to disable unsupported Google Cloud services when working with PHI, to not use a service you cannot disable for PHI, and to not use pre-GA offerings with PHI. The consumer Gemini app on a personal Google account is not covered, because consumer accounts have no BAA with Google. Google AI Studio does not appear on the supported lists we reviewed.
- Shared responsibility. The guide notes that HHS does not offer a HIPAA certification program, and that you must determine whether you are a covered entity or business associate and whether you require a BAA.
Two things follow:
- A clinician at gemini.google.com cannot tell from the URL whether the tab is signed into a personal account or a Workspace account covered by the BAA. Operationally, "Is Gemini HIPAA compliant?" is really "which Google account am I in, and is that account under the Google Cloud BAA?"
- Google's supported-services lists carry their own "last updated" dates and change over time. Verify against the current guide before enabling features for clinical staff.
What a HIPAA-aligned Gemini deployment looks like
A health system, payer, life-sciences org, or digital-health vendor that wants to use Gemini on PHI typically needs to:
- Choose the right product surface. For workforce productivity (drafting clinical notes, summarizing emails, building spreadsheets), Workspace is usually the right path. For application backends, agents, and integrated AI features inside a clinical product, Vertex AI is usually the right path.
- Put the Google Cloud BAA in place. Google's guide describes one Google Cloud BAA that covers both listed Google Cloud services and listed Google Workspace services. Confirm with Google how it applies under your agreement, and confirm that each Gemini surface you use is on the current supported list.
- Configure per the HIPAA Implementation Guide. Disable consumer-facing features that fall outside the BAA scope; enable required logging, encryption, and access controls.
- Restrict workforce access. Block consumer Gemini endpoints (gemini.google.com on personal Google accounts) at the network or identity layer. Force SSO into the BAA-covered Workspace tenant.
- Configure data retention appropriately. Workspace and Vertex AI both support configurable retention; for clinical workloads, set the shortest retention compatible with your records-management obligations.
- Log access to satisfy the Security Rule's audit-control requirement. Workspace exposes Admin audit logs and Drive activity; Vertex AI logs to Cloud Audit Logs and Cloud Monitoring.
- Train the workforce on what may and may not be sent (workforce training is required). A signed BAA does not authorize pasting any PHI, only PHI the workforce member is otherwise permitted to disclose for the specific task.
- Document risk assessment, BAA, configuration, and training records, because the Security Rule requires written policies and records of required actions and assessments, retained for six years (45 CFR 164.316(b)).
Step #4 is the easiest to overlook: the exposure often sits not in the Workspace tenant but in a personal Google account that a staff member opens on a work laptop.
The "shadow Gemini" problem
Even with a properly executed Google Cloud BAA, a common HIPAA exposure in 2026 is workforce members using consumer AI tools — Gemini Advanced, ChatGPT Plus, Claude Pro, Grok, Copilot personal — for clinical and administrative tasks because the personal account is faster and feels less monitored. A Wolters Kluwer Health survey of 518 providers and administrators, fielded by CITE Research in December 2025, found that 40 percent had encountered an unauthorized AI tool at work and nearly 20 percent admitted to using one. One in 10 said they had used an unauthorized tool for a direct patient-care use case, and half pointed to a need for faster workflows.
Two technical controls help materially:
- Device-level redaction. Locke 1.0 detects PHI categories — names, dates, MRNs, diagnoses, drug names, payer IDs — locally and replaces them with one-way placeholders before the prompt leaves the device. Even when staff use a non-sanctioned Gemini Advanced account, the unmasked PHI never crosses the wire. Reversible tokenization is planned shortly after 1.0.
- Egress controls for AI domains. Add gemini.google.com, aistudio.google.com, chat.openai.com, claude.ai, copilot.microsoft.com, and similar destinations to your DLP and CASB rules. Treat them the same way you treat traffic to a personal cloud-storage provider.
Both controls are additive, not substitutes for a BAA. They reduce the size and severity of a leak when policy slips.
Examples: what is and is not allowed
The lines below assume a covered entity has the Google Cloud BAA in place covering Gemini in Workspace. Without that BAA, the answer to every "yes" example is "no, this is an unpermitted disclosure."
| Scenario | Compliant? | Why |
|---|---|---|
| Clinician uses Gemini in Workspace Gmail (BAA, HIPAA features enabled) to draft a reply to a patient referencing their MRN | Yes | Permitted disclosure to a business associate |
| Same clinician opens gemini.google.com on a personal Google account and pastes the same message | No | PHI to a non-BAA account |
| Researcher uses Gemini in Google Sheets to summarize a column of de-identified clinical-trial outcomes | Yes | De-identified data is not PHI |
| Researcher uses Gemini Advanced (consumer) to summarize the same data with patient initials and admission dates | No | Initials + dates remain PHI; non-BAA account |
| Health-tech team calls Gemini on Vertex AI under the GCP BAA from a HIPAA-configured backend | Yes | Permitted disclosure under Google Cloud BAA |
| Front-desk staff drafts an appointment-reminder text in personal Gemini that includes patient name and visit date | No | PHI to a non-BAA account |
| Same staff drafts the message with "[Patient]" placeholders, then fills the name in the EHR | Yes | No PHI sent to Google |
The pattern: keep PHI inside the BAA boundary, or remove PHI before the request leaves the device.
How Gemini compares to ChatGPT and Claude for healthcare
All three vendors can be configured for HIPAA workloads under a BAA. The relevant differences in 2026:
- Training and retention terms. These differ by vendor and plan, and they change. OpenAI says API data is not used for training unless you opt in (API data controls). Anthropic says consumer chats are used to improve its models if you choose to allow it (data-use article). For Google, read the service terms for the specific Workspace or Cloud product you plan to use.
- Cloud breadth. Google's HIPAA coverage spans its productivity suite (Workspace) and its cloud under one Google Cloud BAA. For organizations already standardized on Workspace or Google Cloud, this is procurement-friendly.
- Productivity integration. Gemini's deepest HIPAA-eligible integrations are inside Workspace apps (Gmail, Docs, Drive). For clinicians who already live in those tools, this is a workflow advantage; ChatGPT and Claude live mostly in their own surfaces or via copy-paste.
- Audit and admin controls. Workspace and Vertex AI both expose mature admin and audit-log surfaces. The granularity varies by feature; verify the events your compliance program needs are captured.
The product-level "which is more HIPAA-friendly" question matters less than the organizational question: which BAA do you already hold, where does your workforce already work, and how well can you keep them inside the covered tenancy?
Frequently asked questions
Does signing the Google BAA make Gemini HIPAA compliant?
A BAA is necessary but not sufficient. Compliance requires the BAA plus the Security Rule's administrative, physical, and technical safeguards in your configuration and operation. Google's BAA covers their part; the rest — workforce access controls, audit logs, training, incident response — is yours. Google's own guide notes that HHS does not offer a HIPAA certification program and that compliance is a shared responsibility between you and Google.
Is the free Gemini app HIPAA compliant if I only use it for de-identified data?
If the data is genuinely de-identified under HIPAA's Safe Harbor or Expert Determination methods, it is no longer PHI and HIPAA does not apply. The risk is that "de-identified" is harder than it looks. Safe Harbor requires removing 18 categories of identifiers and that you have no actual knowledge that what remains could identify the person. Expert Determination requires a qualified expert to determine, and document, that the risk of identification is very small. If you cannot document one of the two methods, treat the data as PHI.
What about Gemini in Gmail, Docs, and Drive?
Google's HIPAA guide lists Gemini in Google Workspace and the Gemini app (excluding Gemini in Google Chrome) among the Workspace services included in the Google Cloud BAA, alongside Gmail, Drive (including Docs and Sheets), Meet, and Chat. The guide says not to use pre-GA offerings with PHI, so treat preview or experimental Gemini features as out of scope unless Google's current terms say otherwise.
Is Gemini on Vertex AI the same as the Gemini app for HIPAA purposes?
No. Gemini on Vertex AI runs under the Google Cloud BAA and the Google Cloud HIPAA Implementation Guide; the Gemini app on consumer accounts is governed by Google's consumer terms. Both can use the same model family; the contractual, retention, and infrastructure posture is different. Pick the path your security and procurement teams can support.
Can I use the Gemini API directly without Workspace or Vertex AI?
Google AI Studio and the Gemini Developer API do not appear on the supported-service lists in Google's guide as we reviewed it, and the guide says a service you cannot confirm is supported must not be used for PHI. For HIPAA workloads, use a service on the current Google Cloud BAA list, such as Gemini Enterprise Agent Platform or Generative AI on Agent Platform, and confirm with Google before relying on any other path.
What if my staff already pasted PHI into a personal Gemini account?
Treat it as a presumed breach. Investigate the account's data and retention settings and document the disclosure. Under 45 CFR 164.402, an impermissible disclosure is presumed to be a breach unless you can demonstrate a low probability that the PHI has been compromised, based on a risk assessment of at least four factors: the nature and extent of the PHI involved and the likelihood of re-identification; who received it; whether it was actually acquired or viewed; and how far the risk has been mitigated. Document that assessment. If you cannot show low probability, notify affected individuals without unreasonable delay and no later than 60 days after discovery (45 CFR 164.404), and notify HHS (45 CFR 164.408): at the same time for breaches of 500 or more individuals, otherwise in an annual submission within 60 days after the end of the calendar year.
Your exposure does not depend on leadership knowing. A covered entity is liable for violations by workforce members acting within the scope of their agency, and a breach is treated as discovered when any workforce member other than the person who caused it knows of it, or would have with reasonable diligence.
How does Locke relate to HIPAA when staff use Gemini?
Locke is not a covered entity or business associate. It is a privacy layer that runs entirely on the user's device and never receives PHI or any other data. By detecting PHI categories on-device and replacing them with one-way placeholders before the prompt leaves the device, Locke materially reduces the surface where PHI can leak — including to non-BAA Gemini, ChatGPT, or Claude accounts. It is complementary to a properly executed BAA program, not a substitute for one.
A short HIPAA + Gemini checklist
- Decide which Google products your workforce may use for PHI: Workspace Gemini, Vertex AI, both, neither.
- Put the Google Cloud BAA in place and confirm each Gemini surface you use is on Google's current supported-services list.
- Configure per the current Google HIPAA Implementation Guides (Workspace and/or GCP).
- Block or proxy consumer Gemini endpoints at the network and identity layers.
- Deploy a local-first redaction tool to every browser used by clinical and administrative staff.
- Add AI-bound traffic to your DLP, CASB, and audit-log review.
- Train staff specifically on what may and may not be sent — verify with periodic walk-throughs.
- Re-assess annually; HIPAA-eligible Gemini feature lists change as Google ships new capabilities.
The bottom line
In 2026, "Is Gemini HIPAA compliant?" is the wrong question. The right questions are: which Google product am I using, which BAA covers it, is the configuration aligned with the Security Rule, am I sure my workforce is in the BAA-covered tenancy, and what happens when they aren't? Get those right and Gemini is a useful tool inside a HIPAA program — particularly for organizations already on Workspace or GCP. Skip any one of them and you are running on luck.
Sources and primary references
Reviewed October 1, 2026. Regulations are cited to the eCFR. Vendor statements are cited to the vendor's own documentation as it read on that date; vendors revise these pages, so confirm before relying on them.
- 45 CFR 160.103, definition of business associate — who counts as a business associate, including subcontractors.
- 45 CFR 164.502(e), disclosures to business associates — satisfactory assurance, documented in a written contract.
- 45 CFR 164.504(e), business associate contracts — what the contract must contain.
- 45 CFR 164.312, technical safeguards — access control, audit controls, integrity, authentication, transmission security.
- 45 CFR 164.316, documentation — written policies and six-year retention.
- 45 CFR 164.514, de-identification — Safe Harbor and Expert Determination.
- 45 CFR 164.402 to 164.410, breach notification — the presumption of breach, the four-factor assessment, and notification deadlines.
- 45 CFR 160.402(c), agency liability — liability for workforce members and business associates acting within the scope of agency.
- Wolters Kluwer Health survey press release, January 22, 2026 — 518 healthcare professionals, fielded December 2025.
- Google Cloud, HIPAA compliance on Google Cloud and Google Workspace — the Google Cloud BAA, the supported Workspace and Cloud services, pre-GA guidance, shared responsibility.
- OpenAI, API data controls — API training statement.
- Anthropic, Is my data used for model training? (consumer plans) — when consumer chats may be used to improve models.
Related HIPAA guides
Sonomos publishes this blog and makes Locke, which detects and masks sensitive information on your device before it reaches an AI model. How Locke works
This article is general information about privacy and AI, not legal advice, and reading it does not create a solicitor–client or attorney–client relationship. Regulatory requirements turn on your specific facts and jurisdiction and change over time. Consult qualified counsel before relying on anything here, and check the publication and last-reviewed dates above. We may revise posts, but we do not undertake to keep any post current as the law changes.
Reference pages for this post
Related Articles
The Anthropic DPA Explained: Claude Data Processing Agreement Guide (2026)
Anthropic offers a DPA on Claude Team, Enterprise, and the API — not on Claude.ai Free or Pro. Here is what the Anthropic DPA covers, how it handles EU-US transfers via SCCs and the DPF, what to verify before signing, and where a DPA stops.
Claude EU Data Residency in 2026: What Anthropic Offers, and What European Teams Do Instead
Anthropic does not currently offer EU data residency for Claude — every prompt is processed on US infrastructure and constitutes an EEA-to-US transfer. Here is how that compares to OpenAI, Microsoft, and Google, what the EU-US Data Privacy Framework does and does not cover, and the supplementary-measure stack European teams deploy instead.
Can My Employer See My Claude Chats? What's Visible in 2026 (and What Isn't)
Whether your employer can see your Claude chats depends on the account (personal Claude.ai vs. work-managed Team/Enterprise), the device, and the network. Here is what is visible in each scenario — including Claude Code — and what deleting a chat actually does.