Archive · month
Startup Discovered Materials closed a $9 million seed round from Lightspeed India Partners after completing the Y Combinator accelerator. The company uses AI agents with models from Anthropic to search for materials for cooler and more efficient semiconductors and has released the Material Discovery Bench tool.
According to Živě.cz, Sudolabs and Eldridge aim to bring Sudolabs to unicorn status within a few years. Its revenue is said to be growing by 60 to 70 % year on year. Following the sale of a half stake, the company plans to expand its team in Czechia as well.
New: Eldridge and Sudolabs are targeting unicorn status within a few years.; Revenue at Sudolabs is growing by 60 to 70 % year on year.; Sudolabs has shifted from selling developer hours to implementing AI at large companies.
SemiAnalysis describes TileRT – an engine that compiles the decoding graph into a single persistent kernel on NVIDIA B200 GPUs to achieve 500 tokens/s/user, about 3× more than traditional engines on GB300 NVL72. Already deployed in production at Xiaomi and ZAI.
To accompany the release of the open model Muse Glimmer (30B, Apache 2.0), Meta published an extensive essay by Mark Zuckerberg defending distillation and its open strategy against OpenAI and Anthropic; it also promised to open the weights of Muse Spark 1.2.
New: Mark Zuckerberg published an essay (6000+ words) on AI philosophy, governance and the strategy of Meta; The announcement includes Muse Code, an AI agent for the terminal; Meta strategically positions open-weight models as an alternative to proprietary approaches from OpenAI and Anthropic; The essay criticizes lobbying by competitors and mentions competition from Chinese open-weight models; The new strategy of Meta focuses on differentiation through an open-source approach
Anthropic challenged the legality and process of the US export controls that led it to temporarily suspend access to the Claude Fable 5 and Mythos 5 models in June 2026. In response, it published a jailbreak severity framework and a program for reporting jailbreaks.
New: Export control issued due to concerns about a jailbreak and the model's cyber capabilities; Anthropic challenged the legality of the export control and the lack of a clear process; Anthropic responded by publishing a jailbreak severity framework
According to reporting by The Guardian and Pathfounders, DeepMind is losing its independence and becoming part of Google. Founder Demis Hassabis is reportedly set to leave in the coming months, Koray Kavukcuoglu is taking over operational leadership without the title of CEO, Gemini development is moving to the Bay Area, and Sergey is gaining influence…
Anthropic has added communication between sessions to Claude Code on macOS and Linux – sessions can send messages to each other and share context. On one computer, communication is local; between computers, it passes through Anthropic servers. Administrators can restrict it; it is not available on Bedrock, Google Cloud Agent Platform or…
Backflip AI has released a second AI model that converts 3D scans and mesh files into editable parametric CAD models (extrusions, revolutions) instead of triangular meshes. It runs as an add-in for Autodesk Fusion, with 4 scans free, then $20/month. The company has $30 million in funding.
Since April, Google has been gradually rolling out the Gemini bar in Google Docs to selected users. For personal accounts, it can be disabled through Smart features in Gmail (which also disables proofreading and smart compose); for business Workspace accounts, only the administrator decides. The Bye Bye Gemini extension for Chrome is an alternative.
GitHub added an experimental /worktree command to the Copilot desktop app that creates an isolated workspace for a parallel conversation without disrupting the main work in progress.
Part of an overview of multiple AI topics; only this event has been covered.
GitHub has made the Lite and Balanced levels of detail for GitHub Copilot code review generally available (GA) – they allow the depth of pull request reviews to be chosen based on complexity. Available in the Pro, Pro+, Max, Business and Enterprise plans; admins can set a default level for the entire organization.
OpenAI has released and deployed the model GPT-6 Astra in production – according to the company, it is the most widely deployed model in its lineup and the first to reach the Critical level of cybersecurity capabilities in the Preparedness Framework. The release was preceded by a temporary pause in development due to safety standards.
New: The model GPT-6 Astra has been released and is already deployed in production; It is the most widely deployed model in the portfolio of OpenAI
GitHub has added the totals_by_3rd_party_agent field to the Copilot usage metrics API, breaking down the activity of third-party agents (Claude, Codex, etc.) by individual agent – launch and session counts for enterprise and organization reports.
The research center TReNDS (Georgia State University, Georgia Institute of Technology, Emory University) deployed a system in production that uses CloudWatch, Lambda, Strands Agents SDK and Amazon Bedrock to automatically analyze the causes of errors in applications and send the results to the team through SNS.
Hugging Face recorded 500 errors when downloading files from the CDN in the ap-southeast-1 region (Singapore) during two brief periods starting at 15:14 UTC. The service is running normally, but the underlying cause has not been fully resolved.
According to CEO Brian Chesky, Airbnb used AI to reduce the time from idea to feature release by up to 60 % and increase the number of features released by 80 %. It is testing AI search with a toggle, and an AI support agent resolves 45 % of queries without human intervention.
According to sources familiar with the matter, ByteDance is training an AI model with up to 10 quintillion parameters — three times as many as Kimi K3 from Moonshot. It is in an early stage of pretraining; the estimated size of Mythos 5 from Anthropic is around 8 quintillion parameters.
According to OpenAI, HSP GRUPPE has deployed ChatGPT Enterprise in tax advisory services with the aim of increasing productivity, the quality of work and capacity to serve clients. The source provides no specific results or figures.
České Radiokomunikace is installing five Nvidia DGX B300 servers at the Lužice data center near Hodonín. The investment reaches up to 70 million Czech crowns; one server costs 13–14 million and draws up to 15 kW of power, with liquid cooling.
The Nightrun project released a runtime that allows large language models to run without an operating system. It works on both x86 and Raspberry Pi 5 with at least 4 GB RAM and requires UEFI. The code is under the MIT license and is available on GitHub.
Part of a broader thematic overview; only this AI event has been covered.
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
