Global Outage Sparks Panic as ChatGPT Goes Dark for Five Hours

OpenAI’s ChatGPT platform experienced a widespread disruption on Tuesday, affecting users across multiple continents and sparking renewed concerns over growing reliance on AI systems.

The outage disrupted access to ChatGPT, the Sora text-to-video model, and several APIs integral to OpenAI’s ecosystem.

Initial reports began appearing on Downdetector around 7am UTC, but OpenAI did not publicly acknowledge the incident until two hours later via its status page.

At 9am, the company confirmed “elevated error rates and latency” across its services and stated that it was actively investigating.

By 11am, OpenAI categorised the issue as a “partial outage,” citing problems with 14 APIs, four incidents affecting Sora, and disruptions across 21 ChatGPT components.

Users were met with blank screens or a now-familiar error message:

“Hmm…something seems to have gone wrong.”

The outage affected both free and paid users, with most complaints centered on the web version of ChatGPT.

According to incident metrics, over 1,000 reports came from the UK, with nearly 500 more logged in the US.

Some users also experienced login failures and broken API connections.

While OpenAI has since identified the root cause and is working on a fix, speculation ranges from routine technical bugs to more severe concerns like traffic overload or even a potential cyberattack—though no evidence of the latter has been confirmed.

OpenAI noted:

“We have identified the root cause for the issue causing elevated errors and latency across the listed services. We are working on implementing a mitigation.”

This marks the third major outage of ChatGPT in 2025.

In January, a service interruption hindered DALL·E image generation and slowed GPT-4 response times.

Another significant disruption followed in April, when the platform and its APIs went offline intermittently over two days.

At the time, OpenAI cited issues with its cloud infrastructure, offering limited public explanation.

Performance Dips Following AVM and Translation Launch

OpenAI’s recent service outage coincided with the rollout of key upgrades to its Advanced Voice Mode (AVM) feature in ChatGPT, now available to paid users.

The new functionality, powered by the GPT-4o model, introduces more natural voice interactions and real-time speech translation across multiple languages—without the need to pause between phrases.

These enhancements build on OpenAI’s growing focus on multimodal AI, with GPT-4o capable of both interpreting and generating audio natively.

According to OpenAI’s release notes, users can now instruct ChatGPT to begin translating live conversations, and the assistant will continue translating in real time until told to stop.

This update follows a progression of speech-focused innovations, starting with live translation demos in May 2024, the launch of advanced voice capabilities in September, and the debut of a Realtime API in October that enables developers to embed voice-to-voice functionality into third-party applications.

More voice customisation options were introduced in April 2025.

One X (formerly known as Twitter) user reckoned:

“This changes everything if you’re traveling abroad. ChatGPT can now stay in translation mode no reset needed, just talk. The new voice sounds way more human. More emotion, better pacing. It’s getting scary good.”

However, early adopters have flagged inconsistencies in audio quality.

OpenAI acknowledged that users may experience irregularities in tone, pitch, or delivery depending on the selected voice—particularly in live translation scenarios or when accessibility tools are involved.

OpenAI stated in its release notes:

“We expect to improve audio consistency over time.”

While the core technology is promising, these early hiccups suggest that perfecting real-time voice interaction remains a work in progress.

Research Access Delays Hit Traders and Analysts

Retail investors and crypto analysts increasingly rely on ChatGPT to break down technical whitepapers, analyse tokenomics, and assess smart contract risks.

For this group, even brief outages can delay critical research and decision-making—especially during periods of high market volatility.

Many Web3 platforms also embed GPT-powered bots into Telegram, Discord, and their websites to support user onboarding and community interaction.

When these tools go offline, it disrupts real-time engagement and undermines the user experience, raising questions about the ecosystem’s resilience as AI becomes a core layer of crypto infrastructure.

Affected Users Mocked ChatGPT Outage

As reports of the ChatGPT outage surfaced, users turned to social media platform X to vent their frustrations—and share a laugh. V

Memes and tongue-in-cheek posts quickly filled timelines, with some joking they were incapable of thinking without their AI sidekick or that 90% of their job had evaporated.

Developers quipped about staring at screens, paralysed without code suggestions.

The reaction, equal parts panic and parody, underscored just how embedded AI has become in everyday workflows—and how humor often becomes a reflex when digital tools suddenly go dark.