Fix smart home devices stuck offline after an AT&T outage

Fix status — August 25, 2026

If AT&T internet appears to be back but your smart plugs, bulbs, cameras, or hub still show offline, start with the recovery order, not re-pairing. The useful fix after an AT&T outage is: confirm AT&T service status, restart the AT&T gateway using AT&T’s documented method, wait for the gateway to resync, then restart the rest of the smart-home stack from the network outward. AT&T’s own internet help page says a restarted gateway can take up to 10 minutes to resync with the network, which is exactly the window where smart-home apps often tempt people into doing too much too soon.[1]

The fast recovery order

Run this once, cleanly. Rebooting the same smart plug five times while the gateway is still negotiating with AT&T only adds noise.

1. Confirm whether AT&T still shows an outage for your address.
2. Restart the AT&T gateway: unplug it, remove an internal battery backup if your model has one, wait 20 seconds, plug it back in, and allow up to 10 minutes for resync.
3. If you use a separate router or mesh system, restart it only after the AT&T gateway is stable.
4. Restart your smart-home hub or bridges after the router or mesh is online.
5. Power-cycle remaining offline devices by protocol: Wi-Fi first, then Zigbee/Z-Wave mesh devices, then Thread/Matter checks.
6. Re-pair only the devices that still fail after their controller, network, and protocol path are known-good.

For a more general version of the same network-layer logic, keep Smart Home Not Working After Power Outage? Fix It Step by Step nearby. The AT&T-specific difference here is that the gateway’s service resync is not just another local reboot; it is the first dependency every cloud device, hub, and bridge is waiting on.

First decide whether AT&T is actually back

A phone loading one website is not enough proof that the whole connection is stable. Before touching the smart-home devices, check AT&T’s outage status for your address. AT&T directs customers to check service outages through att.com/outages, and its Smart Home Manager app can also show whether a service outage is affecting the account.[1][2]

If AT&T still reports an outage, stop at the gateway. You can make sure the gateway has power and that cables are seated, but smart-home recovery should wait. Devices that depend on cloud login, DNS, NTP time sync, or a hub connection may look broken while the upstream path is still unstable.

For AT&T Internet Air, make one extra distinction: AT&T tells Internet Air customers to check for wireless outages, not only wired internet outages.[1] The same smart-home restart order can help after service returns, but Air-specific symptoms should be treated more cautiously unless you have verified them on your current All-Fi Hub firmware.

Restart the AT&T gateway the boring, correct way

AT&T’s gateway restart is not complicated, but the wait matters. AT&T says to unplug the power cord from the gateway or modem, remove the internal battery backup if one is installed, wait 20 seconds, plug the battery back in if applicable, reconnect power, and allow up to 10 minutes for the gateway or modem to restart and the broadband light to turn solid green.[1]

That “up to 10 minutes” is where many smart homes get out of order. A router can finish booting before the gateway has a usable WAN path. A hub can finish booting before the router has DNS. A Wi-Fi camera can decide the network is unavailable while the SSID is visible but the internet path is not ready. None of that proves the device is damaged.

If you cannot reach the gateway easily, AT&T also documents a Smart Home Manager remote-restart path for compatible gateways.[3] Use that for the gateway only. Do not use a smart plug to power-cycle the gateway unless you are certain that plug will recover without internet; otherwise you have put the first dependency behind one of the devices that depends on it.

Then rebuild the stack outward

Once the AT&T gateway is stable, move one layer at a time. The order is gateway or modem, then your separate router or mesh, then smart hubs and bridges, then individual devices. The point is not ceremony; it is dependency control. Each layer should come up while the layer below it is already usable.

If your router or mesh creates the Wi-Fi network and the AT&T gateway is in front of it, do not restart both at the same time. Give the gateway its full recovery window first. Then restart the router or mesh so it receives a clean upstream connection instead of caching a half-ready one.

Only after that should the smart-home hub come down and back up. Hubs are often the bottleneck because dozens of devices can look offline when only one controller failed to reconnect. If everything attached to one hub comes back after the hub reboot, the devices were waiting on the hub, not individually broken.

Treat Wi-Fi, Zigbee/Z-Wave, and Matter differently

An app full of offline badges is not one failure. A 2.4 GHz Wi-Fi plug, a Zigbee motion sensor, a Z-Wave lock, and a Matter-over-Thread bulb all have different paths home. Sort the remaining offline devices by protocol before you reset anything.

2.4 GHz Wi-Fi plugs, bulbs, and cameras

For Wi-Fi devices, first confirm your normal 2.4 GHz SSID is present and using the same name and password the device already knows. If your router or mesh came back with a different SSID, guest network, security mode, or band-steering behavior, a cheap plug may look dead when it is only refusing a changed network.

• Power-cycle the offline Wi-Fi device once after the gateway and router are stable.
• Check whether the vendor app shows “offline,” “not connected to Wi-Fi,” or “cannot reach cloud.” Those are different failures.
• If a camera records locally but the app cannot view it remotely, suspect cloud or account reachability before resetting the camera.
• Factory-reset only if the device has lost its Wi-Fi configuration, you changed credentials, or the vendor’s own recovery flow requires setup mode.

When only Wi-Fi devices are offline and the hub-based devices are fine, the likely fault is the Wi-Fi layer: SSID, password, router, mesh node, DHCP, or device cloud reachability. Re-pairing a Zigbee bulb will not help that.

Zigbee and Z-Wave devices behind a hub

If many Zigbee or Z-Wave devices show offline together, look at the hub first. Restart the hub after the AT&T gateway and router are stable. Do not remove devices from the hub, exclude them, or factory-reset them while the hub itself is still recovering.

Mains-powered Zigbee and Z-Wave devices, such as plugs, switches, and in-wall dimmers, usually rebuild their role in the mesh faster than sleepy battery sensors. Battery sensors may not check in until they wake, or until you press their wake button according to the device manual. A lock, sensor, or button that is quiet after the hub reboot is not automatically a failed re-pair.

There is a useful but limited historical clue here. A 2016 SmartThings Community thread described a SmartThings hub on AT&T U-verse staying offline after an internet outage and requiring a reboot; one participant, identifying as a former U-verse tier-2 technician, suggested the hub may stop retrying after an early recovery window.[4] That is a plausible explanation for the symptom many people recognize. It is not a confirmed behavior for current BGW320 or BGW620 gateways, current SmartThings hardware, Hubitat hubs, or Home Assistant boxes.

Use the old thread as a recovery hint, not as a diagnosis label. If the hub comes back only after a hub reboot, write that down as “hub recovery bottleneck.” If one Zigbee bulb stays gone after the hub and nearby powered repeaters are online, then that one device deserves closer attention.

Thread and Matter devices

Matter devices usually should not need a factory reset after an AT&T outage. The Matter fabric, device credentials, and controller relationship are not supposed to disappear because the WAN dropped. The more useful checks are whether the Matter controller, Thread border router, phone, and home Wi-Fi are all back on the same working network.

• For Matter over Wi-Fi, treat the device like a Wi-Fi device first.
• For Matter over Thread, restart the Thread border router or controller after the main network is stable.
• If the device appears in one ecosystem but not another, check controller sync before re-pairing.
• Re-pair only after the controller and border router are confirmed online and the device still cannot be reached.

There are real exceptions. A Home Assistant core issue filed in 2026 describes Matter devices staying unavailable after power cycles, which is enough to justify a cautious exception path, not enough to say Matter devices generally require re-pairing after ISP outages.[5] If your remaining failures are Matter-specific, use a protocol-focused path such as Matter Device Not Showing in Google Home? Start Here before you wipe devices from every app. For broader context on where Matter is heading, Matter 1.6 Brings NFC Setup and Shared Fabrics to Smart Homes is a better read after the house is back online.

When re-pairing is justified

Re-pairing is the expensive move. It can break automations, room assignments, voice assistant links, scenes, and shared access. Use it when the evidence points to one device, not when the whole stack is still unsettled.

A clean rule: re-pair the smallest possible thing, last. If one plug is bad, do not rebuild the whole room. If one hub is stuck, do not reset twenty child devices. If one protocol class is still offline, stay inside that protocol until you can explain the failure.

What the successful step tells you

The step that brings devices back is more useful than the outage itself. If most devices returned after the AT&T gateway restart and full resync window, the failure was gateway or upstream recovery. If phones browsed but the smart home returned only after the separate router restarted, the router or mesh cached the bad moment. If everything behind one hub returned after that hub reboot, the hub was the recovery bottleneck.

If only Wi-Fi devices stayed offline, investigate SSID, password, 2.4 GHz behavior, DHCP, and vendor cloud reachability. If only Zigbee or Z-Wave stayed offline, investigate the hub and mesh. If only Matter or Thread devices stayed offline, investigate the controller and border router before touching factory reset.

Small prevention that actually helps next time

The best prevention is not a pile of spare gadgets. It is making the recovery path boring: keep the gateway, router, and hub on known outlets; label what must start first; and give key hubs or controllers DHCP reservations so their address does not change after a messy restart. If you use a UPS, put the networking core on it before you put decorative devices on it. For weather-related planning and UPS-backed networking, the same principle shows up in Which Smart Home Devices Will Work During a Wind Advisory and Prep Your California Smart Home for a Strong El Niño Winter.

AT&T Internet Backup is a conditional option, not a universal fix. AT&T says Internet Backup requires a BGW320 or BGW620 gateway and an eligible AT&T Wireless unlimited line, is activated through Smart Home Manager, keeps the gateway and connected devices online during a fiber outage, and automatically disables when fiber service returns.[6] That can reduce smart-home disruption for the right household, but it does not remove the need to understand which devices depend on the gateway, router, hub, or cloud.

When the next outage clears, recover the stack in order and note the layer that actually fixed it. That note is what keeps the next hallway full of offline badges from turning into a night of unnecessary resets.

References

1. Get Help With Your Internet, AT&T, link
2. Smart Home Manager (Android), AT&T, link
3. Restart Your AT&T Wi-Fi Gateway or Modem, AT&T, link
4. Hub stays offline after Internet outage, requires reboot (AT&T Uverse), SmartThings Community, 2016, link
5. Matter devices stay unavailable after power cycles, GitHub home-assistant/core issue #161886, link
6. Set Up AT&T Internet Backup, AT&T, link