Hopper · Claude guidance · 1 of 3 Internal guidance · Data handling

Using Claude Desktop at Hopper

Get real work done with Claude — without ever putting patient data at risk.

For: All Hopper & Hopco staff Plan: Claude Team (no BAA) Questions: compliance@hopperhealthcare.com Reviewed: [Month Year] · Owner: [Team]
The one rule

If it could identify a patient, it never goes into Claude.

Everything else — your code, your documents, your planning, your analysis, your vendor research — is fair game and well protected. This page shows you exactly where the line sits, and why business work is safe while patient data is not.

Quick reference

The two lanes

When you’re about to paste, upload, or connect something, ask which lane it’s in. Most work lives comfortably on the left.

Safe to put in Claude
  • Source code & configs — repos, scripts, infrastructure, queries
  • Architecture & technical design — diagrams, trade-offs, decisions
  • Internal docs & strategy — plans, roadmaps, org material
  • De-identified or synthetic data — “Patient A”, SAMPLE-001, relative dates
  • Vendor & market research — dossiers, comparisons, pricing
  • Writing & drafts — emails, memos, docs, summaries
  • Spreadsheets with no patient identifiers — finance, ops, metrics
Never put in Claude
  • A patient’s name alongside anything health-related
  • Member IDs, MRNs, SSNs, DOBs tied to a real person
  • Claims data, EOBs, billing with identifying details
  • Real records from Athena — appointments, charts, encounters
  • Screenshots of the EHR or any screen showing a patient
  • Diagnosis or procedure codes linked to an identifiable person
  • Any combination that could re-identify an individual
Try it

Which lane is this?

Tap an example to see the call. The pattern is always the same: it’s not the topic that’s sensitive, it’s whether the data points back to a real person.

PICK ONE

Tap an example above.The verdict and the reasoning will show up here.

Know the line

What actually counts as PHI

PHI (Protected Health Information) is health, treatment, or payment information tied to someone who can be identified. The trap is the combination: a diagnosis alone is harmless; a diagnosis sitting next to a name, DOB, or member ID is PHI.

Rule of thumb: if a stranger reading it could figure out who the person is, treat it as PHI. Any one of the identifiers below, combined with health or claims information, crosses the line.

Names Dates of birth Addresses Phone & email SSNs Member IDs MRNs Account / claim numbers Health plan IDs Device / serial IDs Photos & face images Biometric data Appointment dates tied to a person Any other unique identifier
Connectors & access

What you can let Claude reach

Claude Desktop can connect to files, tools, and your screen. The question is never “is this tool allowed?” — it’s “does what’s behind it contain patient data?”

Generally fine

Files & folders

OK: code repos, design docs, drafts, finance/ops spreadsheets with no patient identifiers.

Not OK: pointing Claude at folders holding patient exports, downloaded reports, scanned charts, or claims files.

Stop & check

Integrations & connectors

Never connect systems that hold PHI — the EHR / Athena, claims, billing-with-patient-data, or member databases. We have no BAA, so these are off-limits.

OK: GitHub, Linear, internal wikis, and Drive folders you’ve confirmed are PHI-free.

Stop & check

Screen & screenshots

Not OK: sharing your screen or pasting screenshots while anything patient-related is visible — a chart, a worklist, an EHR tab in the background.

Close it first. One visible name or MRN makes the whole image PHI.

Important setting

The 👍 / 👎 rating buttons

Rating a chat is the one action that sends that whole conversation to Anthropic, outside the no-training default. Treat the rating buttons as “send a copy to Anthropic.”

Never rate a chat containing anything sensitive. Admins can switch this off org-wide.

The reassurance

Why business info is safe to use

Your work account runs on Claude Team, which is governed by Anthropic’s Commercial Terms — a different, stronger framework than the consumer plans people use at home.

01

Not used to train models

By default, Anthropic does not train its models on your inputs or outputs under the commercial plans. Source

02

Team sits outside consumer terms

The consumer opt-in training and 5-year retention do not apply to Claude for Work (Team & Enterprise). Source

03

You keep ownership

Anthropic acts as a data processor for the Team plan: Hopper keeps rights to its inputs and owns the outputs.

04

Short retention, no resale

Deleted conversations are purged on a ~30-day cycle, your data isn’t sold, and Claude is ad-free.

So why is PHI still off-limits?

Because business confidentiality is not the same as HIPAA coverage. HIPAA requires a Business Associate Agreement (BAA), and we don’t have one on this plan. The protections above keep your company data safe — they do not make Claude a HIPAA-compliant home for patient data. That’s the whole reason for the one rule. More on commercial data handling

If it happens

If PHI ends up in Claude anyway

Mistakes happen and we assume good faith. What matters is acting quickly and reporting it — there’s always a path forward.

Stop.Don’t send, re-paste, or continue the conversation with the patient data in it.
Delete the chat.Remove the conversation so it leaves your history and the deletion clock starts.
Don’t rate it.Skip the 👍 / 👎 buttons — rating would send a copy out.
Self-report.Email compliance@hopperhealthcare.com with what happened. Reporting early is the right move, not a punishable one.
Common questions

Quick answers

Can my manager or Anthropic read my chats?
Day to day, no. Admins see usage metadata — how much you use Claude, not what you typed. Conversation text only becomes visible through a deliberate, on-demand export run by a Primary Owner. Anthropic staff can’t access your chats by default either.
Can I just use my personal Claude or ChatGPT for work instead?
No. Personal consumer accounts run under consumer terms, where conversations can be retained for up to five years and used for training if the account opted in. Always use your Hopper work account for work.
Is de-identified data really OK?
Yes — if it’s genuinely de-identified. Use placeholders like “Patient A,” MRN: SAMPLE-001, and relative dates (“3 days post-op”) instead of real ones. If you’re reconstructing a real case, make sure no combination of details could point back to the actual person.
What about ICD-10 / CPT codes on their own?
Fine. A code by itself describes a condition or procedure, not a person. You can validate, look up, or reason about codes freely — the line is only crossed when a code is attached to an identifiable individual.
Can I paste an error log or a database query?
Usually yes — but scan it first. Logs and query results sometimes carry patient names, member IDs, or emails in the data rows. Scrub or sample those out before sharing.
This series

More Hopper guidance