Facebook and Instagram Are Down: What We Know About the June 12 Global Outage

TL;DR — Starting around 9:30 AM EDT on Friday, June 12, 2026, users worldwide began reporting that Facebook, Instagram, and Messenger were failing — blank feeds, forced logouts, "unexpected error" and "query error" messages, and login screens rejecting correct passwords. Downdetector showed sharp spikes across the US, Europe, Asia, and the Middle East. Meta spokesperson Andy Stone confirmed the company is "aware people are currently having trouble accessing our services" and "working on it." As of this writing, Meta has not disclosed a root cause or a restoration timeline. This is a developing story — I'll update this post as facts land.

What's Happening

This morning (US time), reports started surging on Downdetector and across X/Twitter that Meta's core consumer products were misbehaving. The symptoms reported so far, compiled from Newsweek, Republic World, and user reports:

• ›Blank or non-loading feeds on Facebook web and mobile
• ›Forced logouts, with some users told their password is incorrect even when it isn't
• ›"Unexpected error" and "query error" messages on Facebook
• ›Instagram feed and login failures, though parts of Instagram remain reachable for some users
• ›Messenger failing to send or load conversations
• ›The web version working for some users while mobile apps fail to refresh entirely

The failure pattern is inconsistent across users — some are fully locked out, others see degraded service — which usually points to a partial backend failure rather than a total network-level event.

Who's Affected

This is global, not regional. Reports so far span:

WhatsApp and Threads have not been widely reported as affected at this time — a notable difference from the October 2021 outage, which took down every Meta property at once.

What Meta Has Said

The only official word so far is from Meta communications lead Andy Stone:

> "We're aware people are currently having trouble accessing our services. We're working on it."

No cause, no ETA. You can watch Meta's own status page for business products at metastatus.com — though historically, consumer-facing incidents show up on Downdetector well before Meta's dashboards turn red.

Is It You or Is It Meta? (It's Meta)

Quick triage if you're debugging on your end:

• ›The issue is server-side. Switching devices, reinstalling the app, or using a VPN will not restore access.
• ›Do not reset your password. The "incorrect password" errors are a symptom of Meta's auth systems failing, not your credentials. Password resets mid-outage can leave your account in a worse state, and outage windows are prime time for phishing emails pretending to be "account recovery."
• ›If your app or website uses "Log in with Facebook" (Meta OAuth), expect elevated login failures for your own users right now. Surface a clear error and offer an alternate login path if you have one.

What Could Cause This? (Honest Speculation, Clearly Labeled)

Meta hasn't said anything about the cause, so anything beyond this point is informed pattern-matching, not reporting.

The symptom profile — auth failures, "query errors," partial availability — looks different from the infamous October 4, 2021 outage, when a bad BGP configuration change withdrew the routes to Meta's own DNS servers and erased facebook.com from the internet for six hours. That one was binary: everything was unreachable for everyone, including Meta's own internal tools and door badges.

Today's failure mode — *some* users logged out, *some* feeds blank, web partially working while apps fail — is more consistent with a bad deploy or configuration push to a core internal service (think auth/session validation or the GraphQL layer that the "query error" message hints at) that's failing for a subset of traffic. Meta runs continuous deployment at enormous scale; the overwhelming majority of historical outages at this layer trace back to a config change interacting badly with production state, not to attacks. There is zero evidence of anything malicious at this point, and outage-day speculation about hacks is almost always wrong.

I'll update this section when Meta publishes a cause or an engineering postmortem.

The Engineering Lesson (Why I'm Covering This)

Every Meta outage is a free masterclass in distributed-systems failure modes, and the recurring theme is worth internalizing for anyone shipping software:

• ›Config changes are deploys. The 2021 outage was a routine BGP audit command. Most catastrophic outages at major platforms (Meta 2021, Cloudflare's regex incident, Fastly 2021) were configuration pushes, not code. If your config pipeline has fewer safeguards than your code pipeline, your config pipeline is your real deploy pipeline.
• ›Blast-radius beats root-cause prevention. You will never prevent every bad change. Staged rollouts, canarying auth changes by region, and the ability to fail open for session validation determine whether a bad push is a blip or a headline.
• ›Don't make a third party your single point of login. If "Log in with Facebook" is your only auth path, your uptime is capped at Meta's. Offer email/passkey fallback.
• ›Status communication is a product feature. A single two-sentence statement from a spokesperson, hours into a global outage, is the gap your own status page should aim to beat.

Current Status

• ›Outage began: ~9:30 AM EDT / 7:00 PM IST, June 12, 2026
• ›Services affected: Facebook, Instagram, Messenger (WhatsApp and Threads largely spared so far)
• ›Cause: Not disclosed
• ›Meta's statement: Aware and working on it
• ›Historical precedent: Similar Meta outages have typically resolved within one to three hours

*This is a developing story. I'll update this post as Meta restores service or publishes a cause.*

---

*Harsh Rastogi is an AI Product Engineer at Modelia, building production Generative AI systems for fashion commerce. He writes about AI systems, developer tooling, and production engineering at harshrastogi.tech.*

Written by Harsh Rastogi — AI Product Engineer leading AI product direction at Modelia. Connect with me on LinkedIn for more on Shopify, Generative AI, agentic systems, and production engineering.