Sign in
Edition of 10:00 CETThursday, 11 June 2026
287 outlets · 16 languages77 briefings today
Thursday, 11 June 2026 · Edition of 06:00 CET

Google’s AI push stumbles as new translation tool launches amidst global outage

The rollout of Gemini 3.5 Live Translate, designed to enable seamless voice-to-voice communication across 70 languages, was marred by a widespread service disruption affecting millions of users, while the company quietly extended its browser assistant into new markets.

Technology5 outlets3 languages3 min readUpd. 09:31

Google’s ambition to eliminate language barriers suffered an immediate blow on Tuesday when the launch of its most advanced speech-to-speech translation model was upstaged by a global outage of the very AI platform meant to power it. The new Gemini 3.5 Live Translate, built on the Gemini 3 Pro architecture, promises real-time, interruption-free conversation in over 70 languages directly within Google Translate, Meet, and through an API for developers. It is a technology that, in the tech giant’s telling, should make awkward pauses in multilingual exchanges a relic of the past.

That vision, however, collided with reality as users from Jakarta to Madrid reported being met with error messages when attempting to use the core Gemini service. Indonesian news outlets chronicled the disruption across Android, iPhone, web browsers, macOS, and even integrated Google Workspace tools, with the platform returning codes 1099 and 1076 alongside the familiar “Something went wrong”. Spanish-language analysts noted that error 1099 blocked new conversations entirely, while 1076 pointed to a breakdown in server communication, suggesting a infrastructure failure rather than a superficial glitch. The timing could hardly have been more awkward for a company betting its future on ubiquitous, always-available AI.

Even as engineers scrambled to restore service, Google pressed ahead with its geographical expansion of Gemini’s presence. In a move largely unnoticed amid the outage headlines, the company confirmed that its AI assistant in Chrome is now available to users in a swathe of countries across Latin America, Africa, the Middle East, and Asia—including Indonesia, Thailand, Vietnam, and Laos—while conspicuously continuing to exclude the European Union. Viewed from Delhi or São Paulo, this expansion signals a determined push into emerging digital markets, but the day’s technical difficulties left many first-time users confronting a platform that seemed less than ready.

The dual narrative of breakthrough and breakdown exposes the precarious balancing act facing Google as it races to embed AI into every corner of its ecosystem. The Live Translate feature is technically a marvel, yet it relies on the same backend that faltered globally, raising questions about what happens when a supposedly seamless conversation is cut off mid-sentence. Analysts in London note that the incident underlines a broader industry tension: the faster large language models are shipped to users, the larger the risk that critical infrastructure lags behind.

Looking ahead, the episode is unlikely to slow Google’s momentum, but it will sharpen scrutiny on the resilience of its AI layer. The company has now tied its translation, meetings, and browsing services to the same Gemini core, creating a single point of vulnerability that, when it goes dark, leaves users across continents staring at the same cryptic error codes. For a tool designed to bring the world closer together, Tuesday’s outage briefly achieved the opposite effect.

This story appeared in

5 sources · 3 languages · 24h window

La RazónJun 11, 02:31
TribunnewsJun 11, 04:34
Los AndesJun 11, 03:30
TechNewsJun 11, 08:32
Antara NewsJun 11, 05:33