Google Gemini Outage
Google's Gemini AI Chatbot Experiences Widespread Global Outage, Disrupting User Access
Key Points
Google VP Josh Woodward publicly acknowledged the outage on X, writing: “Gemini is currently experiencing an outage. We're on it and will get everything back up ASAP. Some of the fixes are in, the rest coming very soon.”
Google’s Workspace dashboard/status communication gave a specific start time—"our engineering team continues to investigate the issue"—and said the disruption began at 3:26 a.m. PT (6:26 a.m. ET), apologizing to those affected.
One workaround reported by users was that “immediately resubmitting a failed prompt often pushes it through,” even while others were seeing 1076/1099 failures across web and mobile.
Downdetector’s breakdown suggested the mobile experience was most impacted: the Gemini app accounted for 58% of reports, the website for 35%, and Google Home/automation workflows for only 5%.
A Google Support product-expert (as quoted by Android Police) described error 1076 as “a browser-level conflict or a temporary communication glitch with Gemini” and recommended troubleshooting like switching models, doing a hard refresh, or using Gemini in an incognito browser tab for repeated 1099 errors.
Google's Gemini AI went dark for millions of users on Wednesday, June 10, in an outage that lasted over seven hours. Starting at 6:26 a.m. ET, users across the US, UK, Europe, and parts of Asia found themselves locked out — unable to start new chats, load history, or get any response at all, according to 9to5Google.
Error codes 1076 and 1099 flooded user screens across web, iOS, Android, and macOS. Reports on Downdetector peaked at over 1,600 simultaneous complaints, with thousands more rolling in throughout the morning. Google VP Josh Woodward finally broke silence at 10:31 a.m. PT, writing on X: "Gemini is currently experiencing an outage. We're on it and will get everything back up ASAP."
Status Page Said 'All Clear' While Users Were Locked Out
The outage started at 3:26 a.m. PT, but Google's public status page kept showing "all systems operational" for nearly three hours after user reports spiked. Complaints on Downdetector began a near-vertical climb at 6:11 a.m. ET — long before any official word. Users on Reddit expressed shock that thousands were unable to work while the status dashboard stayed green, according to TechRadar.
By 10:31 a.m. PT, Woodward posted his acknowledgment. Minutes earlier, Google's Workspace Dashboard quietly noted that mitigations were applied and the "majority of users" were recovering. The company finally resolved the incident at 5:30 p.m. PT — 7 hours and 4 minutes after it began — citing a "performance issue in our backend database fleet," per the Google Workspace Status Dashboard.
One Backend Database Broke Gemini Across Four Continents
The same two error codes — 1076 and 1099 — showed up in user reports from the US, UK, Europe, and Asia. In a TechRadar user poll, 89% of affected users saw one of those two errors. The other 11% saw prompts silently bounce back into the text box with no response. The fact that identical failures hit every region at once points to a single, central backend component, not a regional network glitch.
Error 1076 was described as a "handshake or initial connection timeout" — a breakdown between your browser and Google's servers. Not all models failed equally. Gemini Flash and Gemini Pro were hit hardest, while Flash Lite still answered some requests intermittently, TechTimes reported. Some users found that resubmitting a failed prompt immediately would sometimes push it through to a working server.
Mobile Users Hit Hardest, Workspace Workflows Disrupted
Downdetector data showed the mobile app drove 58% of all reports, with the website accounting for 35% and Google Home or automation integrations making up the remaining 5%. Gemini's role inside Google Workspace — reading emails, pulling from Docs, accessing Calendar — was specifically broken because the outage hit the "tools catalog" retrieval layer. That's the system Gemini uses to call outside apps, according to Dataconomy.
A Google Product Expert advised users to start a new chat to escape "poisoned" sessions and try an incognito browser tab for repeated 1099 errors, as quoted by Android Police. US reports peaked at 765 and UK reports at 440 during the worst hours, per TechTimes.
Outage Exposes Risk of Betting Everything on One AI Provider
The disruption hit just weeks after Google's I/O 2026 conference, where the company unveiled the Gemini 3.5 family and pushed hard on "agentic" AI — tools that act on your behalf inside Gmail, Docs, and Drive. Tech analysts told WebProNews that outages like this are no longer mere app hiccups. They are "business-process failures" as AI moves into the daily rhythm of work.
Experts are now recommending that businesses build "failover" strategies — routing to Anthropic's Claude or OpenAI's GPT if one provider goes down — to stay operational, per Dataconomy. Google classified the event as a "Service Disruption" rather than a full outage, but Workspace SLA terms may still entitle enterprise customers to service credits for the 7-hour gap.
Sentiment
Mostly NegativePublishers
19
Articles
6
Reach
25
