AI Flash · outage history
We keep operational incidents separate from the regular news stream. Today's outages are featured prominently on the homepage, while older ones remain here as a searchable history.
Continuously updated
112 items in the selected feed
How to read evidence statuses ↓
Requests to the Claude Fable 5.1 model are experiencing an elevated error rate. According to the operational report dated 25 October at 08:17 UTC, the cause has been identified and work on a fix is ongoing.
The Claude Status page reported increased error rates for requests to five Claude models. The notice, published on 25 October at 06:14 UTC, states that the incident is under investigation and another update will follow as soon as possible.
According to the Claude Status page, the cause of increased error rates for requests to unspecified Claude models was identified on 24 October at 05:24 UTC. Work on a fix was ongoing, and resolution of the incident had not yet been announced.
On 24 October 2026, xAI reported degraded model performance in the api.x.ai API. It announced an investigation at 00:03:44 UTC and a resolution at 00:04:21 UTC. The available record does not state the cause or the extent of the impact.
According to GitHub, Grok 4.6 and Grok 4.7 have reduced availability in Copilot, including within the IDE. The report attributes the disruption to an issue at the model provider. The company recommends selecting another model or the Auto option.
According to the Claude Status report, the cause of elevated error rates affecting Claude Fable 5.1, Sonnet 5.5 and Haiku 5.5, as well as fast mode for Opus 5.5, has been identified. Work on a fix continues.
xAI recorded an outage affecting Grok models in Grok Bot on 23 October 2026 at 23:39 UTC. After deploying a fix, it announced on 24 October at 00:20 UTC that the incident had been resolved. Approximately 41 minutes elapsed between the initial report and closure.
According to an update from Claude Status on 22 October at 07:28 UTC, a fix has been deployed for connector issues in claude.ai and Claude Cowork. Service recovery is being monitored; full resolution of the incident has not yet been confirmed.
OpenAI has deployed mitigation measures following ChatGPT issues in India, Japan and Korea and API errors. The incident is in Monitoring status; degraded performance continues to be reported for the Conversations and Responses components.
The Claude Status page reports an elevated error rate for requests to the Claude Fable 5 model. The update from 21 October at 19:49 UTC lists the incident as under investigation.
The status report indicates an elevated error rate for the FedRAMP API service for Astra and Sol. The FedRAMP component is experiencing a partial outage, the incident status is Identified, and measures to mitigate the impact are being prepared, according to the report.
According to a status report from OpenAI, all affected services have fully recovered from file download errors. Affected components included Responses, Image Generation and file services.
OpenAI announced that it had deployed a mitigation measure for widespread degradation of functionality. The incident status is Monitoring; the company has not yet confirmed that service has been fully restored.
In one notice, OpenAI reports delays in publishing cost data for organizations through Costs API and a status of Identified; a second notice reports a status of Resolved. Neither the chronological order of the notices nor the current status can be determined from the available information. According to the notice with the Identified status, neither use of the API nor use of ChatGPT is affected.
OpenAI announced that elevated error rates in Work Mode have been mitigated and is monitoring recovery. The status page continues to list Codex in ChatGPT Desktop and ChatGPT Work as experiencing degraded performance.
According to OpenAI, the GPT-6.1 Sol model is experiencing degraded performance in Codex tools and ChatGPT Work. The incident is under investigation and affects the web interface, CLI, the VS Code extension and Codex API.
According to OpenAI, the ability of workspace administrators to view and manage plugin availability and settings is being restored. The incident remains in the Monitoring state; the Connectors/Apps component is marked as operational.
OpenAI has identified an issue that prevents the ChatGPT app for Android from displaying some conversation content. A fix is being rolled out gradually in the latest version of the app; the company points to the web version as a temporary option.
According to OpenAI, some users in the APAC region are encountering errors in ChatGPT Work, conversations and GPTs. All three components are experiencing degraded performance, and the incident is under investigation.
According to OpenAI Status, all services affected by an increased number of errors from the GPT-5.6 Instant model in FedRAMP workspaces have fully recovered. The available excerpt does not state the duration of the incident or its cause.
According to a service status notice, some organizations were mistakenly suspended because they were believed to have reached their spending limit. Claude API, Claude.ai, Claude Code, Claude Security and Claude Cowork rejected their requests. The issue has been resolved.
The Dot feature in Codex may fail to create new threads. OpenAI is investigating the issue; according to the status page, the dots component is experiencing reduced performance.
The operator has mitigated errors in loading usage data on platform.claude.com and via Admin API. Work on a full fix continues; data for approximately the last 30 minutes may appear on the portal with a delay. According to the announcement, API requests are not affected.
OpenAI released a fixed version of the desktop app for Linux, macOS and Windows. It addresses a problem with creating new ChatGPT Work threads. The incident is marked as resolved; affected users should update the app.
AI Radar monitors Czech and international sources every day, looking for changes that truly deserve attention.
The result is a quick overview of what has actually changed in the AI world, rather than another stream of articles.
Use the CS/EN switch to read the same Radar in Czech or English. English content is published after its translation has been checked, so new and older items may appear later.
Each row represents one event — not one article. At a glance, you can see its significance, credibility and main point.
Illustrative example, not a current news item.
Agent mode is available to all paying users
Until now, the mode was available only on the highest plan; it is now available on all paid tiers without a waitlist.
The detail page contains a fuller summary, its significance and original sources. Practical impact appears in the detail and the For individuals and For businesses views. An AI Flash item reaches the main selection only after it has been expanded and meets the publication rules.
We first summarize each event in the same way for everyone. Based on those same facts, we then explain what the change means for your own use and what it could mean for how a company operates.
Today’s briefing is the same for everyone. Pages For individuals and For businesses can be found in the main navigation — they select only events relevant to the given use case.
Reset history · Outage history · Subscribe in a feed reader (RSS) · Methodology and label explanations
