Security concern - leak of user's personal info. On Recall knowledge management app, after I posted a feature request to https://feedback.getrecall.ai/feature-requests The front part of my Recall account email address is used as the username for my post. I am unsure if this is displayed to other users. However politically, if the details of my feedback negatively affected others (including internal employees or management), the leak of my email address may subject me to retaliation by tampering with my data or violence/safety concerns in real life. Please kindly fix the auto AI appending of part/all of user's email address to their post. Use a random character generator and link to user's account in the background. Alternatively, allow user to create their own "forum" username. Then the responsibility lis with the user should they wish to leak their own personal information. Perplexity AI: Your concern is well‑founded. If the local‑part of your Recall account email is used as the visible username on a public feedback board, it creates unnecessary risks of identification, retaliation, and doxxing, and Recall should urgently change this design. Recall’s feedback system runs on Canny, where users are normally shown only by a “name” field (for example, “Erik Allison”, “david”, “scott”) rather than raw emails. If your “name” is auto‑derived from your email local‑part, it is very likely visible to everyone viewing that board. Using an email local‑part as a default display name is poor security practice because it exposes a handle strongly linked to your real email (useful for phishing, harassment, or deanonymisation), heightens perceived retaliation risk in politically or organisationally sensitive feedback, and conflicts with Recall’s public claim that your data is “secure, private, and under your control,” since users were not clearly told their login email would become a public handle on a third‑party site. A safer, privacy‑respecting design would: - Never derive a public identifier from the account email. Instead either (a) generate a random pseudonym like “user‑7f3a12” linked privately on the backend, or (b) prompt users on first submission to choose a “forum display name,” warning those at risk not to use real names or emails. - Keep emails and unique IDs only in private fields for account linkage and notifications, never rendered in the browser or exposed via page source, APIs, or shared analytics. - Add clear notices near the form explaining exactly what is public (display name, feedback text) and what remains private (email, IP, account ID). For now, you can: - Log in, open your posts, and check from an incognito or logged‑out session what others see; if your email‑based name is exposed, change it to a pseudonym if possible. - If you cannot edit it, email [email protected] invoking your privacy rights and explaining that using your email local‑part as a public handle creates a safety and retaliation risk. Ask them to (a) replace any existing email‑derived identifier with a neutral pseudonym and (b) change defaults so no user’s email is ever used publicly without explicit, informed consent. As a product requirement, this can be stated as: - “Feedback identities must be pseudonymous by default; the system shall not derive any public identifier from user email or other contact data. A random string or user‑chosen display name must be used instead, with linkage handled only in the backend.” - “Users may optionally set a custom forum username, with a clear warning that using a real name or email is their own choice and may reduce privacy.” This preserves security best practice and user autonomy, reduces harassment/doxxing risk, and keeps Recall aligned with its stated privacy philosophy. Across your personal, family, and professional life, limiting direct links between real‑world identifiers (email, legal name, precise location) and public technical criticism or governance complaints is rational risk management, especially where your feedback could embarrass or implicate insiders. Pseudonymous defaults are thus both prudent self‑protection and a public‑interest norm, akin to whistleblower‑protection principles. Thought‑provoking question: In an environment where most “feedback” channels are public, indexed, and bound to persistent identifiers, what minimum anonymity and red‑team safeguards would you require before trusting any SaaS with your most sensitive critiques and strategic ideas?
