According to Anthropic, the new safeguards reduced the volume of internal reasoning tokens extracted from Claude models by 90–99 %. A classifier detects extraction attempts, and flagged accounts are automatically blocked. The company acknowledges that legitimate users are also affected.
According to records reviewed by the editorial team at WIRED, the Think Big committee paid approximately 20 million dollars to Lantern Production Consultants. Campaign Legal Center filed a complaint with the FEC over the alleged concealment of the actual recipients of the payments; the Leading the Future network denies violating the law.
The analysis examines the benefits and risks of AI models with available weights. According to the article, Anthropic describes the GLM-5.3 model as a significant cyberattack risk. The available excerpt presents a dispute over safety, broader access and providers’ commercial interests.
According to the article, Alza.cz has four configurations of AI servers from Gigabyte in stock, with two units of each. The price is 15.5 to 16.8 million Czech korunas. Each server contains eight Nvidia B300 Blackwell Ultra SXM accelerators and 2 160 GB of HBM3E memory.
SimpleClosure brokers sales of code and work archives for AI training. Keelan Miskell received approximately 25 000 dollars for his code. Surge AI uses archives to create environments for training agents; an advantage over alternatives has not yet been demonstrated through comparative tests.
Startup Comparables.ai, which uses AI to find companies suitable for acquisition, raised 6 million dollars in a seed round, approximately 130 million Kč. Czech fund Purple Ventures contributed 10 million Kč. The claim that search time can be reduced from months to minutes comes from the startup.
Mirakl uses fine-tuned and open models running on GPUs and inference software from NVIDIA. According to NVIDIA, it has added over 70 million products this way and reduced the time required to onboard seller catalogs to less than two hours.
Part of an overview of multiple AI topics; only this event has been covered.
In an essay for Bulletin of the Atomic Scientists, Deb Roy advocates holding specific people accountable for public statements created using AI. He points out that artificially generated comments on public policies are harder to identify.
Liquid AI has made LFM2.5-24B-A2B available for local agents. It has 24 billion parameters, with 2 billion active per token, and a context window of 131 072 tokens. According to the company, it generates 184 tokens per second on Apple M5 Max hardware.
Gemini 4 Argon became generally available in Gemini API on 13 October 2026 under the identifier gemini-4-argon. According to the provider, it is designed for complex programming, agentic tasks, mathematics and science.
Version 0.2 of the llm-meta-ai plugin adds support for Muse Wonder. Simon Willison also shows its outputs at six reasoning effort levels, from minimal to max.
According to GitHub, enterprise administrators can link GitHub Copilot activity to specific users without collecting prompts and responses. Five content categories can be controlled separately. Identity collection and collection of all content categories are disabled by default.
According to a Bloomberg report, AI models are to generate all the code and direct the creation of graphics for an upcoming shooter from a 1.5 billion dollar collaboration between Epic Games and Walt Disney Company. The release planned for March 2027 is reportedly being postponed indefinitely.
According to the available excerpt, Gurmeet Saran develops infrastructure at Anthropic to understand how Claude behaves. His work covers data reliability and privacy protection, which he says must be part of the architecture from the outset.
Elon Musk posted a screenshot of a purchase using Grok Bot that included addresses, an email address and an order number. Users claimed they had canceled or changed the order through the chatbot from Lego. Musk denied that the order had been canceled; Lego did not confirm any changes.
AWS has published a guide on calling GPT-6 models from source regions in Asia through global cross-Region inference in Amazon Bedrock. The client passes a global profile identifier; the service selects the destination region for processing.
GitHub describes how to provide GitHub Copilot with project rules: store shared instructions in the repository, define rules for specific files using applyTo, and describe recurring workflows or specialized tasks using agent skills and custom agents.
The application web-edit 0.1a0 lets users view and edit files through a web interface. The author created it to work with Codex Remote from his phone so he could change login credentials without passing them to an LLM. It is designed to run over a Tailscale network.
AWS published a guide to text generation and scoring predefined options on Amazon SageMaker AI. It uses llama.cpp and the SemIf adapter; the example includes deployment on both CPU and GPU, along with accuracy and latency measurements.
The guide from AWS shows how to deploy a single LLM on Amazon EKS across two GPU nodes: one processes the input, and the other generates the response. Ray LLM Deep Learning Containers connect both phases behind a shared interface compatible with OpenAI API.
Honeyjar offers an AI tool in beta for preparing announcements, finding journalists and monitoring media coverage. Companies and PR agencies are testing it. Users remain responsible for sending drafts and publishing posts.
Blueprints AI plans to unveil the second generation of its software for construction drawings at TechCrunch Disrupt. According to the founders, it is meant to take over repetitive drafting. The available excerpt does not specify new features or a release date.
Roman Horáček presented the planned iSoud system for summarizing court filings and preparing materials to support legal arguments. It could be operational within a few months. According to Horáček, it could free up over 229 thousand hours of judges' work annually; the article does not report measurements behind this estimate.
The commentary uses the film Annihilation to explore how sci-fi shapes ideas about the future of AI. It questions the predominance of apocalyptic scenarios and raises the question of whose values determine the direction of development. The available excerpt does not substantiate any specific positive impacts.
AI Radar monitors Czech and international sources every day, looking for changes that truly deserve attention.
MonitorsOfficial AI company blogs, specialist media, and research sources.
Selects and combinesFilters out information noise and combines articles about the same change into a single event.
Summarizes and explainsExplains significant events in English: what happened, why it matters and where the information comes from.
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.
Everything you need to navigate the AI world
Today’s briefingAn overview of news from the past 24 hours and a main story when available. Beside them is Live · AI Flash, followed by research and links to other Radar sections. On mobile, the blocks appear one below another.
AI FlashAn ongoing feed of brief updates with an evidence status. Links lead to a Radar detail page when one is ready, otherwise to the original source. You can also find reset and outage histories here.
Practical applicationsWhat new tools and features can do, what you can try and what their actual impact could be.
Model selectionModel comparison by type of work, capabilities, price and speed.
Research and archiveA separate research overview, topic search and older events by date.
One event, everything that matters
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.
Importance: ▮▮▮ majorOpenAIModels✓ 6
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.
▮▮▮ major · ▮▮ important · ▮ we're tracking = how significant the change is✓ 6 = six independent publishers, not the number of articles or feeds✓ official = a clear release, law or incident is substantiated by the relevant authority1 source = no independent confirmation yetbold = who is behind the changegray text = a brief summary of what happened
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.
The same news, two practical uses
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.
For individualsWhat you can use or try, how the change can help you at work and what to watch out for.
For businessesWhat impact the change could have on processes, costs, risks and other business decisions.
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.