Screen Privacy for Researchers and Academics

You're at the conference hotel between sessions, finishing the slides for tomorrow's talk. The data table on slide 12 shows the subgroup analysis you haven't yet published. The person across the lobby is a friendly rival who works on the same problem at a different lab. They saw you sitting there and walked over to say hi. They saw slide 12 first. Now your scoop is on borrowed time.

Researchers carry data that's regulated, irreproducible, and competitively valuable, and they work in the most porous environments imaginable: shared labs, library carrels, conference halls.

What's on your screen

A researcher's open files in a given week:

• Subject data, often with identifiers, recruitment IDs, demographic info
• Unpublished findings, including the killer figure from the in-progress paper
• IRB protocols, including the specific consent language and approved procedures
• Grant proposals with budget detail and specific aims
• Peer review materials, including the manuscript you're confidentially reviewing
• Co-author drafts with track changes and candid comments
• Datasets that violate data use agreements if redistributed (CMS, restricted-use NCES, ICPSR datasets)
• Statistical code with hardcoded paths revealing the data source
• Email threads with collaborators about candid disagreements

The peer review case is particularly delicate. Confidential review materials are open on the laptop in environments where the authors of the manuscript might literally walk past.

Where shoulder surfing happens for researchers

Conferences are the main risk: hotel lobbies, conference rooms, hallway gatherings between sessions. The university library, where the carrels are open and other students walk past. The lab common area, where lab mates and visiting collaborators move around your bench. The shared office space junior faculty often get. The campus café. The plane to a conference. The visiting talk where you're at the front of the room and someone behind you can see your laptop during Q&A.

The lab common area case is the most everyday. Lab mates from competing groups, rotation students, visiting scholars, all move through the same physical space.

The unique risks

The exposure is regulated, reputational, and competitive.

IRB violations: institutional review boards require strict confidentiality of subject data. Casual disclosure, even visual, can trigger an IRB investigation that suspends the protocol.

Subject re-identification risk: even when names are masked, combinations of demographic variables visible on a screen can re-identify a participant. HIPAA Safe Harbor identifies eighteen specific identifiers, and several can be inferred from quasi-identifiers.

Retraction risk: data integrity questions raised by improper disclosure have triggered retractions and corrections.

Scooping: in competitive fields, an unpublished finding seen by a rival lab can compress the publication timeline or split priority.

Grant confidentiality: NSF and NIH proposals contain confidential strategy. Reviewers see them under non-disclosure. Casual disclosure violates the review process.

Data use agreement breaches: restricted-use datasets (HRS, NLSY, hospital EHRs) carry data use agreements that explicitly prohibit unauthorized observation.

The figure on your screen is the result of three years of work. The colleague who saw it from across the lobby published it first not by being smarter, but by being faster on a Tuesday afternoon.

Practical methods

Research environments are unusually open and unusually informal. The methods have to fit that culture.

Privacy filter on the laptop. Even one rated at 60 degrees handles most library carrel and lobby cases.

Encrypt subject data at rest so the file you're working with is in a mounted encrypted volume that closes if the laptop sleeps. FileVault plus a separate Veracrypt volume is a common setup.

Blur faces in screenshots and figures before pasting them into talks. Most reidentification cases come from raw image data in slides.

Use participant codes, not names, even in your in-progress analysis script. The grep result that pops up in a public laptop view should reveal codes.

Never leave the conference laptop unattended, even for two minutes. Library carrels have laptop locks for a reason.

Confidential peer review files only at home or in the office. Never in the conference hotel where the authors might be present.

Hide the file tree in your IDE in public. VS Code's "Zen mode" is excellent for this.

See privacy filters vs software for the tradeoff details.

Where camera-based detection fits in

The conference and the library carrel are where camera-based detection earns its keep. Peeker runs locally on the Mac, watches for faces behind you using the webcam, and shows a small preview in the corner. Nothing leaves the device. Nothing is recorded.

For a researcher in a conference lobby or a library carrel, that quiet awareness layer can catch the rival lab member walking up to say hi before they finish reading slide 12.

FAQ for researchers

Is Peeker IRB-approvable for use with subject data on a research laptop? IRB approval applies to protocols, not to individual apps. Peeker doesn't process or store subject data because it only handles camera frames locally. It can be one element of a secure research workstation. Confirm with your IRB and IT.

Will it work alongside RStudio, Stata, or MATLAB? Yes. It runs in the background with no interaction with statistical software.

My lab uses university-managed laptops. Can I install it? Depends on the university's policy. Send your IT team the link, the signed app, and the privacy policy. Most local-only camera utilities get approved quickly.

Wrap

Research depends on data integrity, subject trust, and priority of discovery. All three can be eroded by a screen seen in the wrong place. Peeker is $5/year, and quietly covers the awareness layer.