April 2026 AI News Archive

Small models also found the vulnerabilities that Mythos found

Article URL: Comments URL: Points: 359 # Comments: 106

Google's Gemma 4 puts free agentic AI on your phone and no data ever leaves the device

Google's new open-source model, Gemma 4, processes text, images, and audio completely on-device. Using agent skills, the AI can independently tap into tools like Wikipedia or interactive maps; no cloud required. The art…

AI models would rather guess than ask for help, researchers find

ProactiveBench tests whether multimodal language models ask users for help when visual information is missing. Out of 22 models tested, almost none ask for what they need, but a simple reinforcement learning approach hi…

Claude Code's new Ultraplan feature moves task planning to the cloud

Anthropic's new Ultraplan feature for Claude Code moves task planning to the cloud. Claude works out the plan in the browser while the terminal stays free for other work. The article Claude Code's new Ultraplan feature…

Anthropic Wants to Run Your AI Agents For You — Here’s the Real Cost (It’s Not $0.08/hr)

Two days ago, Anthropic’s announcement hit 2 million views in two hours. Within the same hour, a developer on HackerNews wrote what might… Continue reading on Towards AI »

Coreweave signs multi-year cloud deal with Anthropic to power Claude

Coreweave has signed a multi-year cloud deal with AI startup Anthropic. The article Coreweave signs multi-year cloud deal with Anthropic to power Claude appeared first on The Decoder .

Meta has a competitive AI model but loses its open-source identity

The open-source AI movement has never lacked for options. Mistral, Falcon, and a growing field of open-weight models have been available to developers for years. But when Meta threw its weight behind Llama, something sh…

ChatGPT finally offers $100/month Pro plan

OpenAI announced on Thursday something that power users have been asking for: a $100/month plan. Previously, subscriptions jumped from $20 to $200 per month.

Meta AI app climbs to No. 5 on the App Store after Muse Spark launch

The app was ranking No. 57 on the App Store just before Meta AI's new model launched. Now it's No. 5 — and rising.

Is Anthropic limiting the release of Mythos to protect the internet — or Anthropic?

Anthropic said this week that it limited the release of its newest model, dubbed Mythos, because it is too capable of finding security exploits in software relied upon by users around the world. Are real cybersecurity c…

Anthropic keeps new AI model private after it finds thousands of external vulnerabilities

Anthropic’s most capable AI model has already found thousands of AI cybersecurity vulnerabilities across every major operating system and web browser. The company’s response was not to release it, but to quietly hand it…

Meta's new model is Muse Spark, and meta.ai chat has some interesting tools

Meta announced Muse Spark today, their first model release since Llama 4 almost exactly a year ago . It's hosted, not open weights, and the API is currently "a private API preview to select users", but you can try it ou…

ALTK‑Evolve: On‑the‑Job Learning for AI Agents

GLM-5.1: Towards Long-Horizon Tasks

GLM-5.1: Towards Long-Horizon Tasks Chinese AI lab Z.ai's latest model is a giant 754B parameter 1.51TB (on Hugging Face ) MIT-licensed monster - the same size as their previous GLM-5 release, and sharing the same paper…

Welcome Gemma 4: Frontier multimodal intelligence on device

Falcon Perception