Audacity 4.0 Released with UI Overhaul and Refreshed Clip Editing Model
The open-source audio editing software Audacity has released version 4.0, featuring a reconstructed UI and the introduction of a new clip editing model.
Latest
Microsoft Office Head Warns Against Low-Quality AI-Generated Content
Ryan Roslansky, head of Microsoft's Office division, discusses the risk of a vicious cycle where low-quality AI-generated documents undermine productivity.
Anthropic Releases Blueprints for Building Commerce Agents
Anthropic has released blueprints for rapidly building commerce agents in sectors such as retail and travel, featuring capabilities like catalog search and inventory management.
AWS Releases AWS Cloud Quest 2.0, a New Version of Its Learning Game
Amazon Web Services has launched AWS Cloud Quest 2.0, a new version of its online game that allows users to learn AWS technologies within a 3D open world.
NYC Department of Education Announces One-Year Ban on Generative AI for Students from Age 2 through Middle School
The Mayor of New York City and the NYC Department of Education have announced a one-year suspension of generative AI use for public school students from age 2 through 8th grade for the 2026-2027 academic year.
Google Reveals Details of Mantis, an Open-Source Framework Automating Vulnerability Discovery to Remediation
On September 2, 2026, Google released details regarding Mantis, an open-source framework that leverages AI to discover, reproduce, and fix software vulnerabilities.
US Department of Justice Files Brief Supporting OpenAI, Arguing AI Training is Fair Use
In the copyright infringement lawsuit filed by The New York Times against OpenAI and Microsoft, the US Department of Justice has submitted a brief to the court stating that the training of AI models constitutes fair use under copyright law.
Adobe Launches Adobe for Slack, Enabling Access to Photoshop and Other Tools Directly Within Slack
On September 2, Adobe launched Adobe for Slack, allowing users to operate tools such as Photoshop and Firefly within Slack. By providing instructions in natural language via Slackbot, users can generate images and videos based on the context of their conversations.
Anthropic Open-Sources Agents for AI Development in Online Stores
Anthropic has released "commerce agents," a developer-focused tool for integrating AI with product recommendation and analysis capabilities into online stores, targeting sectors such as retail, travel, and telecommunications.
Google Announces Gemini 3.8 Flash and Security-Focused Cyber Model
On September 2, 2026, Google announced Gemini 3.8 Flash, featuring enhanced software development and inference capabilities, along with Gemini 3.8 Flash Cyber, designed to assist in vulnerability discovery and remediation.
Efforts by Sumitomo Chemical and LINE Yahoo to Achieve Company-wide Adoption of Generative AI
Through cases such as Sumitomo Chemical, LINE Yahoo, and DIP, specific strategies have been presented to ensure generative AI becomes an established part of business operations rather than just a passing trend.
Fraudulent Schemes Mimicking CEOs' Voices via AI Result in 4.5 Billion Yen in Damages
A former Metropolitan Police Department official has pointed out that total damages from fraud involving the imitation of CEOs' voices using AI have reached 4.5 billion yen through 2026.
Meta Releases AI Model Muse Spark 1.3, Outperforming Latest GPT and Claude in Coding Benchmarks
Meta has released Muse Spark 1.3, a new AI model with enhanced capabilities for coding tasks. The company announced that the model recorded performance exceeding GPT-5.6 Sol and Claude Opus 5 in benchmark tests.
codex rust-v0.153.0 Released — Vim Mode Undo/Redo Support and Plugin CLI Enhancements
OpenAI has released rust-v0.153.0 of the local LLM development tool "codex." The primary changes include the addition of undo/redo functionality in Vim mode and remote marketplace integration via the plugin CLI.
Meta Announces Muse Spark 1.3, Enhancing Coding and Agent Performance
Meta announced the AI model Muse Spark 1.3 on September 2 (local time), aimed at improving agent tasks and coding performance, and has begun offering it through developer services.
Bushiroad to Establish AI Subsidiary Bushiroad Mirai in December
Bushiroad has announced the establishment of a subsidiary, Bushiroad Mirai, in December 2026 to promote the use of AI for operational efficiency.
World No. 1 Shin Jin-seo Defeats Go AI KataGo with a Two-Stone Handicap
South Korean Go Grandmaster Shin Jin-seo has secured a victory against the AI engine KataGo in an official series played under a two-stone handicap condition.
Proliferation of open-weight Models May Lead to More Efficient Cyberattacks
The availability of high-performance open-weight models is lowering the barrier to entry for cyberattacks. Experts emphasize the importance of AI-driven defenses and strategies based on the assumption of breach.
Attempting to Reproduce a JavaScript-based Watercolor Generation Model Using TRL and OpenEnv
A method for reproducing a model that paints watercolors using JavaScript, proposed by Surya Narreddi, has been published on the Hugging Face blog. The developer utilized TRL and OpenEnv, providing everything from the reference dataset to the pre-trained model openly.
Method for Improving Structured Output in 350M-Scale Models Using GRPO Released
Hugging Face has released a method to improve structured output compliance for the small LFM2.5-350M model using Group Relative Policy Optimization (GRPO).