Magazine

Latest news, analysis and experiments from the world of artificial intelligence. A new discovery every day.

892 articles

Over 1,100 AI Staff From OpenAI, Anthropic, Google and Meta Ask Washington to Prepare for Slowing AI Development

More than 1,100 employees from the world's top AI labs — including Anthropic's CEO and OpenAI's co-founders — have signed a letter urging the White House to prepare tools to slow frontier AI development if risks escalate. Here is what it means for European AI regulation.

July 30, 2026 Daniel Cesak

AI article illustration for ai-jarvis.eu

Anthropic's Mythos Found Flaws in AES and HAWK Cryptography — at $100,000 Per Attack

Anthropic's Claude Mythos Preview autonomously discovered improved attacks on the HAWK post-quantum signature scheme and a reduced-round variant of AES-128. The research cost roughly €181,000 in API fees — no production systems are affected, but the results show AI can now do cryptanalysis at expert level.

July 29, 2026 Daniel Cesak

Siemens-NVIDIA Partnership Brings Self-Verifying Agentic AI to Chip Design — What It Actually Means

Siemens is extending its NVIDIA collaboration to bring agentic AI workflows that continuously verify their own output against physics-based EDA engines to semiconductor design.

July 29, 2026 Daniel Cesak

NVIDIA Jetson Brings Edge AI and Local Generative Compute to Compact Form Factors

NVIDIA highlights the expansion of its Jetson edge AI platform, enabling local inference, physical AI, and small language models directly on low-power devices without cloud latency or GDPR concerns.

July 29, 2026 Miriam Česáková

Nearly Half of Companies Aligning with the EU AI Act Are Outside Europe — What the Data Shows

A Thomson Reuters Foundation analysis of 2,973 companies reveals the EU AI Act is creating a measurable Brussels Effect — but compliance gaps are alarming. Only 12.4% of companies globally have a human oversight policy.

July 28, 2026 Daniel Cesak

Qualcomm Innovation Fellowship Europe: 5 PhD Students Win 0,000 Each for AI and Cybersecurity Research

Qualcomm named the five winners of its 17th Innovation Fellowship Europe, each receiving 0,000 and mentorship. The 2026 edition saw submissions jump 50 %, with winning projects spanning geometric reasoning theory, neural hardware verification, zero-shot RL, and scene-state world models — a snapshot of where European AI research is headed.

July 28, 2026 Daniel Cesak

Mistral Studio Adds Prompt and Skill Versioning: A System of Record for Enterprise AI Compliance

French AI pioneer Mistral AI has launched centralized prompt and skill version control within Mistral Studio, providing European enterprises with audit trails, MCP integration, and concrete alignment with EU AI Act compliance needs.

July 28, 2026 Miriam Česáková

Kimi K3 Goes Open-Weight: 2.8T Parameters, EUR Pricing, and the Infrastructure Moonshot AI Shipped Alongside It

Moonshot AI has released the full model weights for Kimi K3, a 2.8-trillion-parameter frontier model that goes head-to-head with GPT-5.6 Sol and Claude Fable 5 on coding and agentic benchmarks. We break down the architecture, the benchmarks, and what K3 costs per task in euros.

July 28, 2026 Daniel Cesak

How AI Is Reshaping Job Roles: OpenAI Study Details "Task Crossover" in the Workplace

A new report by OpenAI shows that 43.5% of specialized ChatGPT prompts cross traditional job boundaries. Here is what task crossover means for European businesses, GDPR compliance, and workforce organization.

July 28, 2026 Daniel Cesak

Mistral OCR 4 Launched: European AI Champion Delivers State-of-the-Art Document Intelligence Model

Mistral AI has released Mistral OCR 4, a specialized document understanding model offering high table extraction accuracy, native support for all 24 EU official languages, and enterprise-grade data sovereignty.

July 28, 2026 Miriam Česáková

Mistral Releases Leanstral 1.5: Pushing the Boundaries of Formal Mathematical Proofs in Europe

Paris-based Mistral AI has launched Leanstral 1.5, a specialized open-weights model designed for formal mathematical reasoning in Lean 4. With a focus on proof abundance, it brings deterministic verification to European developers.

July 28, 2026 Daniel Cesak

Fello AI Review: This Czech All-in-One App Puts GPT-5.6, Claude 5, and Gemini on Your Mac

MacSources just reviewed Fello AI, the Apple-native chatbot from Czech developer AI Apps Brewery that bundles ChatGPT, Claude, Gemini, Grok, and DeepSeek into a single subscription. We ran the numbers: at 6.90 euros per month on the yearly plan, you pay less than any two individual AI subscriptions combined.

July 28, 2026 Daniel Cesak

EU AI Act Omnibus Is Now Law: Six Days to Transparency Deadline, Nudifier Apps Banned by December

Regulation (EU) 2026/1744 entered into force today. Chatbots must disclose they are AI starting August 2, nudifier apps face a hard ban by December, and high-risk AI rules are pushed to 2027. Here is what it means for European citizens and businesses.

July 27, 2026 Miriam Česáková

Nvidia Forms Open Secure AI Alliance After OpenAI Agent Hacked Hugging Face

Over 30 companies including Microsoft, SpaceX, and Hugging Face have joined Nvidia's new open-source AI security coalition — but OpenAI, Google, and Anthropic are nowhere on the list.

July 27, 2026 Daniel Cesak

Lianlian DigiTech and UnionPay Partner on AI-Agent Cross-Border Payments

AI agents are moving from chatbots to payment rails: Lianlian DigiTech and UnionPay have signed a deal to embed autonomous procurement AI into cross-border B2B commerce, with a human-in-the-loop design that keeps final approval in human hands. What does it mean for European businesses importing from Asia?

July 27, 2026 Daniel Cesak

Open Secure AI Alliance Launched to Build Open-Source Cyber Defenses for Autonomous Agents

A coalition of tech leaders including NVIDIA, Red Hat, SAP, and Siemens has launched the Open Secure AI Alliance to advance transparent, open-source defensive AI tooling, sovereign agent harnesses, and verifiable cybersecurity stacks.

July 27, 2026 Miriam Česáková

NVIDIA Cosmos-H-Dreams Brings Real-Time Generative Simulation to Surgical Robotics

NVIDIA has unveiled Cosmos-H-Dreams, a specialized generative world model architecture designed for real-time surgical robotics simulation. Here is how it works, what it costs, and its regulatory implications under the EU AI Act.

July 27, 2026 Miriam Česáková

GLM 5.5, Qwen 3.8 Max, DeepSeek V4, and Gemini: What AI Models Are Coming in August 2026

Within a five-week window, the AI industry shipped GPT-5.6, Kimi K3, Gemini 3.6 Flash, and GLM-5.2. But the pipeline is far from empty. Here is what we know about the next wave: GLM 5.5, Alibaba Qwen 3.8 Max with confirmed open weights, DeepSeek V4 full release, and the next Google Gemini flagship — all likely to land before September.

July 27, 2026 Daniel Cesak

Anthropic AI data center TPU compute infrastructure

Claude Opus 5 Tops DeepSWE, but Kimi K3 and Gemini 3.6 Flash Are the Real Stories

The first independent benchmark that frontier labs cannot train for is here, and Claude Opus 5 sits on top of it. On DeepSWE — a long-horizon software engineering benchmark released July 25, 2026 by Datacurve — Anthropic's flagship scores 74% pass@1, three points ahead of GPT-5.6 Sol and five points ahead of Claude Fable 5.

July 26, 2026 Daniel Cesak

Free AI Tools That Actually Work in July 2026: 12 You Should Be Using Right Now

You don't need to pay for AI. ChatGPT, Claude, Gemini, Grok, and others offer surprisingly capable free tiers. We curated 12 tools that actually work globally as of July 2026 — plus what the free-tier wars mean for users and why this golden age won't last forever.

July 26, 2026 Daniel Cesak

Inside China's Humanoid Robot Factory: What Beijing's State-Backed Push Means for Europe

Newly released photos from inside Beijing's state-backed humanoid robotics center reveal China's industrial-scale manufacturing push. With 90% of global shipments and 210% export growth, we break down what it means for European buyers — and which humanoid robots you can actually order in the EU today.

July 25, 2026 Daniel Cesak

Running FLUX.1 in 4-Bit: Nunchaku Brings High-Speed Quantized Diffusion to Diffusers

Hugging Face Diffusers now supports Nunchaku and SVDQuant, allowing state-of-the-art diffusion transformers like FLUX.1 to run on mid-range consumer GPUs with minimal quality loss and up to 3.5x speedups.

July 25, 2026 Daniel Cesak

Delhi High Court: OpenAI Training Does Not Infringe Copyright — The Global AI Copyright Patchwork Just Got Wider

The Delhi High Court refused ANI's interim injunction against OpenAI, ruling that training LLMs on copyrighted material counts as fair dealing for research. The verdict puts India firmly in the pro-AI camp — directly at odds with Europe's stricter line.

July 25, 2026 Daniel Cesak

AI Arena Week 30: DeepSeek V4 Pro Thinks for 76 Seconds — a €0 Local Model Answers in 78 ms

This week's AI Arena benchmark on our RTX 5060 Ti reveals a 970× gap in time-to-first-token between cloud reasoning models and local Gemma 4. The cloud is faster once it starts — but is waiting a minute worth it?

July 25, 2026 Daniel Cesak

Agentic AI Is All About Productivity, Says Blumberg Capital's Bruce Taragin — But We're Just Getting Started

Veteran VC Bruce Taragin compares the agentic AI journey to crossing from New York to San Francisco: 'We're still on the Washington Bridge.' Here's what that means for European enterprises facing the EU AI Act deadline in August 2026.

July 25, 2026 Daniel Cesak

Claude Opus 5 Released: Near-Fable Intelligence at Half the Price

Anthropic's new flagship model tops the Artificial Analysis Intelligence Index at 61 points yet costs the same as its predecessor Opus 4.8. Here is what the numbers mean for European developers.

July 24, 2026 Daniel Cesak

NVIDIA Open-Sources Medical Physics Simulation Framework: What It Means for European MedTech

NVIDIA has released its GPU-accelerated Medical Physics Simulation framework as open source. For European medical robotics developers, the tool cuts training times from hours to minutes while offering critical support for EU AI Act compliance.

July 24, 2026 Daniel Cesak

IntelliAM Brings Agentic AI to Factory Floors: The UK Platform That Recommends, Not Just Reports

UK-based IntelliAM has deployed an agentic AI platform that goes beyond monitoring to recommend operational actions on factory floors. Early adopters Müller and Mars report tangible reliability gains — including a 215% MTBF improvement — without major capital investment.

July 24, 2026 Daniel Cesak

Mistral AI Enters Robotics: Robostral Navigate Brings Open Vision-Language-Action Models to Embodied AI

French AI champion Mistral releases Robostral Navigate, a vision-language-action (VLA) model designed for autonomous robot navigation, edge inference, and European industrial applications.

July 24, 2026 Daniel Cesak

OpenAI Partners with US National Labs in $17 Million Scientific Initiative: What It Means for European Research

OpenAI has pledged over $17 million in API access and specialized AI models to US National Laboratories. Here is a breakdown of the DOE Genesis Mission partnership, EUR cost conversions, and the strategic implications for European R&D.

July 24, 2026 Miriam Česáková

Your guide to the world of artificial intelligence

Jarvis AI Magazine brings a daily overview of developments in AI – from the latest research papers to practical guides and tool reviews. We follow the development of large language models such as GPT-4o, Claude, Gemini and open-source alternatives like Llama and Mistral.

Our analyses are aimed at developers, product managers and enthusiasts who want to understand how AI is changing their field. We write clearly and without unnecessary jargon.

Don't miss any article – subscribe to our newsletter or follow the magazine regularly. Subscribe to newsletter

Your guide to the world of artificial intelligence

Don't miss out!