Social Media Privacy 6-8 min read · May 06, 2026

X (formerly Twitter) Privacy & Security Guide 2026

X is fast-moving and highly public. Geo-tagged posts, reply patterns, and follower lists are all OSINT goldmines.

Start with discoverability rather than the padlock: disable Discoverability by email and by phone number, because those two turn a leaked address into your handle in a single step. Then turn on Protect your posts, switch Photo tagging off entirely, and work back through old posts that name a workplace or a neighbourhood.

Not ready yet? Run a free breach check on this email

We’ll check it against 13.1B+ leaked records right now — no account needed. Continuous monitoring & alerts are part of Protection.

Key steps to lock down X (formerly Twitter) in 2026

Profile → Settings and privacy → Privacy and safety.
Turn on Protect your posts.
Set Who can message you to People you follow.
Enable sensitive media marking and hide sensitive content.
Turn off Photo tagging entirely.
Disable Discoverability by email and by phone number.
Audit and delete old tweets that reference location, employer, or family.
Enable two-factor authentication via an authenticator app.

X (formerly Twitter) privacy controls illustration

Quick checklist

Profile visibility: Private or friends-only
Search engine indexing: Off
Location sharing: Off
Two-factor authentication: Enabled (authenticator app, not SMS)
Data partner sharing / personalized ads: Off
Linked apps + sessions: Audited and revoked where unfamiliar

Why these settings still aren't enough

Even with every X (formerly Twitter) setting locked down, your data still leaks through three channels these settings can't reach: historical exposures already in breach corpora, third-party scrapers that mirrored your old public data, and people-search aggregators that re-list your details every time you remove them. That's where continuous external monitoring becomes essential.

How Warden extends your X (formerly Twitter) privacy

Warden by GalaxyWarden monitors X usernames and linked data across breach corpora, including the 2022 Twitter API leak (5.4M records).

Run a free Warden scan to see exactly what is exposed about you across every platform — not just X (formerly Twitter).

You’ve used today’s 5 free articles · resets tomorrow

Create a free account to keep reading

The rest of this article comes straight back — free, no card, nothing you’ve already read goes away. Then see whether your own data is exposed too: the scan on this site is free with or without an account, and the account is what keeps the result after the emailed link expires.

Free reader account

No card, no charge. You read the blog without a limit — every article, no daily count — and it saves your exposure-scan results so you can open them again after the emailed link expires. We won’t email you anything else unless you tick the box asking us to. The next step shows what the paid plans add — free stays the first option on it.

Create a free account →

Free accounts are for personal use. If you’re reading for a company, that’s Signals (terms).

Reading this for a company?

GalaxyWarden Signals watches the organisations you name and emails you when one of them actually discloses — SEC 8-K Item 1.05 filings and corroborated breach records, with a JSON API. It monitors companies, not people: it does not scan you and does not remove your personal data.

See GalaxyWarden Signals →

Or make a free reader account first

Already have an account? Log in

Already exposed?

You can’t unleak data. You can take away what it’s worth.

A leaked record is where it starts, not where it ends. What turns it into your front door is the look-up sites publishing your address beside your name — and those are what an AI reads when somebody asks about you. The free scan shows you both. We write to 580 companies.

See what is exposed about you — free scan →

Keep Reading

The Persona-to-Identity Link: How Doxxers Actually Connect Your Handle to Your Real Name

Auditing Your Own Doxx Surface: A 30-Minute Self-Check for Streamers and Creators

Why a One-Time Scan Isn't Enough: The Case for Continuous Persona Monitoring