AI News

Inspired by Anthropic's strategy, OpenAI is unveiling a specialized model: GPT-5.4-Cyber. This cutting-edge tool is laser-focused on advanced coding and security analysis, but for now, access is extremely restricted.

Chinese tech giant AGIBOT just dropped a bombshell at APC 2026 in Shanghai! They unveiled five cutting-edge robotic platforms, eight powerful new AI models, and announced the deployment of their 10,000th robot. It's official: the humanoid robotics industry is moving beyond experiments and into real-world productivity.

Alibaba just dropped Qwen3.6-35B-A3B, an open-source AI model under the Apache 2.0 license! It scored an impressive 73.4% on SWE-bench Verified, blowing past Gemma 4 and nearly matching top proprietary models. Plus, it blazes past 120 tokens per second on an RTX 4090!

AI's monumental leap: No longer just simple text responses, but autonomous systems that plan and execute tasks. We're diving into GPT-5.3-Codex, Claude Opus 4.6, and the fierce tech battle for transparency.

Agent AI is no longer just about conversation. Get ready for the era of autonomous systems that don't just assist in cybersecurity, but actively make decisions. How do we harness their power and ensure our safety?

Beyond chatbots, autonomous agents are here. Agentic AI isn't just a buzzword anymore – it's a real, transformative power that's reshaping how we code, conduct business, and even how we pay for software.

AI agents are unlocking autonomous decision-making, but they're also unleashing a wave of new cyber threats. Discover how Galileo and the upcoming OWASP 2026 standards are setting the secure boundaries for these autonomous systems.

In the tech world, a silent war is raging for dominance in logical reasoning. A 'laden knight's tour' test has just exposed crucial differences between Claude, Gemini, and GPT, fundamentally reshaping how developers and businesses operate.

The legal battle for AI accountability is taking a dramatic new turn. OpenAI is facing lawsuits after ChatGPT was implicated in aiding stalkers and contributing to tragic outcomes. What are the dangers of "overly helpful" AI models?

Chinese startup AGIBOT just produced its ten-thousandth humanoid robot and partnered with NCS and HZTC for deployment in Singapore and across Asia. How will this game-changing breakthrough impact Europe and Czech companies?