Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesMicrosoft’s Bing outage on May 23, 2024 was a dependency incident, not a shutdown of the entire internet. Bing’s homepage and parts of Bing Search failed for some users, Microsoft Copilot was affected, and services that used Bing results—including DuckDuckGo, ChatGPT’s internet search and Ecosia—also reported problems. Some direct Bing search URLs continued to work while Bing.com showed a blank page or an HTTP 429 error.
What happened
Reports placed the incident on May 23, 2024. Search Engine Land reported Bing Search problems beginning at about 1:30 a.m. Eastern Time. BleepingComputer reported the disruption at approximately 3 a.m. Eastern Time and said users appeared to be affected mainly in Asia and Europe. Those times and regional observations come from separate contemporaneous reports, so they should not be treated as a complete outage map.
As an Amazon Associate I earn from qualifying purchases.
The most visible failure was Bing.com itself. Users could receive a blank homepage or an HTTP 429 (“Too Many Requests”) response. BleepingComputer also reported that opening a direct Bing search URL could still return results for some users. That difference matters: an unavailable homepage did not necessarily mean every Bing Search request was down.
Free tools Windows power users keep installed
One-click scans. No signup required.
Which products and services were affected?
| Service or layer | Reported effect | What the evidence establishes |
|---|---|---|
| Bing homepage | Blank page or HTTP 429 error | Observed by BleepingComputer; regional scope was not fully established. |
| Bing Search | Broad failures were reported | Search Engine Land reported recovery after about five hours. |
| Direct Bing search URLs | Some continued to work | Reported by BleepingComputer, meaning homepage and search-endpoint behavior could differ. |
| Microsoft Copilot | Unavailable or degraded | Contemporaneous reports linked its problems to the Bing-related incident; its recovery window was longer than the Bing Search window in some coverage. |
| DuckDuckGo, ChatGPT internet search and Ecosia | Search-related features reported problems | Search Engine Land identified these as downstream services that rely on Bing search in relevant features; their outage durations were not identical or fully established. |
Why a Bing failure reached other search and AI products
Many products do not crawl the web independently for every query. They call a search provider or use that provider’s index as one component of an answer pipeline. When Bing Search or a related internal component cannot answer those requests, a dependent product may show no results, return an error, or disable only its web-search feature while the rest of the product remains available.
That dependency model explains why the incident looked different across products. A Bing homepage request, a direct search request, Copilot’s retrieval step and a third-party search API call are separate paths. They can fail at different times and recover independently.
Microsoft’s response and the reported cause
Microsoft said it was investigating. In an update relayed by BleepingComputer, the company said, “We’re working to isolate the cause of the issue.” A later update said Microsoft was “transitioning requests to alternate service components to expedite service recovery.”
BleepingComputer’s May 24 follow-up reported that Microsoft attributed resolution to a problem with an internal service-to-service call. The available coverage does not identify the specific component, call, deployment or configuration change involved, so the incident should not be described as having a more precise root cause than that.
How long was Bing down?
There is no single duration that applies to every product mentioned in the reports:
- Bing Search: Search Engine Land said it was back after approximately five hours, measured from its reported early-morning start.
- Copilot and related Bing services: BleepingComputer described impacts lasting longer than 24 hours for Bing/Copilot-related services in its follow-up coverage.
- Third-party services: The reports identified downstream problems but did not establish one shared start or end time for DuckDuckGo, ChatGPT’s internet search and Ecosia.
These figures describe different service scopes. Combining them into one “Bing outage length” would incorrectly imply that every endpoint and dependent product recovered simultaneously.
What users could do during the incident
- Try a direct Bing search URL if the homepage was blank or returned HTTP 429; some users reported that this path still worked.
- Use another search provider that was not relying on the affected Bing path.
- If only Copilot’s web-connected answers failed, treat that as a product or dependency issue rather than proof that all Copilot functions were unavailable.
- Retry later instead of repeatedly refreshing a page returning HTTP 429, which can indicate request throttling or an overloaded service.
What this outage demonstrates
The event shows why “search availability” is not a single switch. A provider can have a working index while its homepage, routing layer or internal service call is failing. Conversely, a provider outage can appear as a failure in an AI assistant or a competing search engine that consumes its results. Evaluating an outage therefore requires identifying the failed layer, the affected products, the observed geography and each service’s recovery time.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
