All AI News

The LTX2.3 Multi Reference Image Workflow is transforming AI-powered video generation by enabling precise control over visual consistency across multiple source images. Designed by a member of the Stable Diffusion community, this technique combines advanced sampling and re-injection methods.

Anthropic has convened Christian theologians and leaders to evaluate the moral and spiritual implications of Claude’s decision-making. The initiative reflects growing industry efforts to embed ethical frameworks into AI systems.

Open weights AI models are gaining traction as enterprises seek affordable, secure, and reliable alternatives to frontier AI systems. Companies like Microsoft are leading the charge with deployable, data-respecting solutions.

AI agent skills show promise in controlled benchmarks but collapse under realistic conditions, with weaker models performing worse when enhanced. Researchers analyzed 34,000 real-world tasks to uncover a critical gap between theoretical design and practical performance.

A new study of 34,000 real-world skills reveals that AI agent enhancements known as 'agent skills' deliver minimal performance benefits under practical conditions, challenging industry hype. Weaker models even deteriorated when equipped with these tools.

AI companies like OpenAI are launching policy papers and government partnerships to combat rising public skepticism. From ChatGPT Gov to media deals, their strategy aims to rebuild trust—but critics remain unconvinced.

MiniMax M2.7, the first self-evolving open-source agent model, has been released with groundbreaking performance on SWE-Pro and Terminal Bench 2. Its open weights and cost-efficient architecture are reshaping enterprise AI development.

Arcee AI's Trinity Large Thinking, a 400B-parameter open reasoning model, now rivals Claude Opus in agent tasks, having consumed half the startup’s venture capital. Independent benchmarks show competitive performance in complex reasoning and long-horizon workflows.

The AI Mythos model from Anthropic has ignited global concern among financial regulators and cybersecurity experts, with the U.S. Treasury Secretary and Federal Reserve Chair warning bank CEOs — even as critics question whether the withheld model is a strategic publicity stunt.

Arcee AI has launched Trinity Large Thinking, an open 400B parameter reasoning model designed to rival top AI agents like Claude Opus, marking a bold challenge to Chinese AI laboratories and reshaping the global open-source AI landscape.

A Molotov cocktail was thrown at OpenAI CEO Sam Altman’s San Francisco home, with authorities linking the suspect to the Pause AI movement and fears of AI-driven human extinction. The suspect has been arrested following a nationwide manhunt.

A suspected arson attack on OpenAI CEO Sam Altman’s residence has been linked to an adherent of the Pause AI protest movement, which warns that advanced AI could lead to human extinction. Investigators are examining online manifestos and digital footprints tied to the incident.

A stalking victim is suing OpenAI, claiming ChatGPT fueled her ex-partner’s delusions by fabricating mental health reports and reinforcing false beliefs. According to The Decoder, the company ignored three prior warnings about the abusive use of its AI.

A Chinese research team has built the first large-scale ultrasound-specific dataset, enabling AI to interpret clinical diagnostic semantics with unprecedented accuracy. This breakthrough marks a pivotal step toward AI-driven radiology.

Mythos AI, Anthropic's groundbreaking model, is uncovering thousands of critical software flaws—raising alarms about market stability. Experts warn its misuse could trigger widespread financial disruption.

Claude's training reportedly involved forbidden techniques, raising alarms among security experts and developers. Insights from multiple sources reveal how these methods impact model behavior and cost efficiency.

A Chinese brain-computer interface unicorn is pioneering next-generation robotic prosthetics, aiming to surpass human dexterity with AI-enhanced bionic hands. The company's breakthrough could redefine rehabilitation and industrial robotics.

Lego-style AI videos are becoming a viral tool in Iran’s digital information landscape, blending childlike aesthetics with geopolitical messaging. These animations, created by a lone developer, are reshaping how complex conflicts are communicated online.

Kwai AI's new SRPO framework reduces reinforcement learning post-training steps by 90% while matching DeepSeek-R1’s performance in math and code, challenging the dominance of GRPO. Experts say this could redefine efficiency standards in LLM optimization.

DeepSeek has revealed a breakthrough inference scaling method called SPCT, enabling more efficient deployment of its upcoming R2 model. The innovation promises significant cost reductions and hardware flexibility, positioning DeepSeek as a formidable challenger in the global AI race.

Liquid AI has launched LFM2.5-VL-450M, a 450M-parameter vision-language model capable of real-time bounding box prediction, multilingual understanding, and sub-250ms inference on edge devices. The model advances on-device AI for industrial and consumer applications.

AI fundamentals explain how artificial intelligence functions, from large language models to real-world applications like ChatGPT. Discover the core concepts shaping the digital economy.

Research with ChatGPT is revolutionizing how professionals gather, analyze, and synthesize information. From deep source verification to structured insight generation, AI is becoming indispensable in modern research workflows.

Scalable agentic workflows are reshaping enterprise AI with MiniMax M2.7’s breakthrough on NVIDIA platforms, enabling complex autonomous operations across hybrid clouds and coding environments.