An analysis of 20 AI chatbots found that 17 sent third-party tracking code.
Orthrus, which uses a Diffusion module to boost token generation speed by up to 7.8x while maintaining the accuracy of existing Transformers, has been unveiled.
OpenAI has piloted a ChatGPT personal finance feature for US Pro users.
VS Code has revealed its Copilot harness optimization and the VSC-Bench evaluation system.
GitHub reviewed 3,535 PRs and resolved 68% of them, revealing the design and operating principles of its accessibility agent.
Image-blaster, an open-source tool that integrates the generation of 3D environments, meshes, and sound effects from a single image, has been released.
DwarfStar 4, a Metal-based local inference engine, has been released supporting DeepSeek V4 Flash.
Claude Code 2.1.142 adds file delivery and agent-related updates.
NextGenU expanded voice narration and localization for global health education with ElevenLabs.
Anthropic's Claude Mythos helped develop a macOS exploit in 5 days, and Apple has fixed the flaw.
Runway is challenging Google by expanding from video generation into world models.
Amid the spread of AI tools, GitHub has tightened the quality standards and responsibility boundaries of its bug bounty program.
Microsoft will end use of Claude Code by the end of June and switch to its own GitHub Copilot CLI.
Turso has discontinued its bug bounty program due to a surge in low-quality AI-generated code (slop).
Anthropic launched 'Claude for Small Business,' which includes 8 major tools and 15 agent workflows.
Amazon extended the Chinchilla scaling law to present a framework that speeds up LLMs without losing accuracy.
We examine the background of Anthropic's limited release of Mythos from both the security and cost angles.
NextGenU expanded voice production for global health education with ElevenLabs.
Survey results examining AI usage and risks among more than 7,000 web developers have been released.
Thoth inspects and tests repos and folders to turn them into reusable Custom Tools.
LLM-based code scanning tools have begun uncovering security vulnerabilities in open source projects en masse
ByteDance has released checkpoints for Cola-DLM, a continuous latent diffusion language model.
A template error in the Qwen3.5 Vision model has been fixed in vLLM.
Anthropic's AI model Claude found a security vulnerability in macOS and took control of the system in just 5 days.