Liquid AI released DSpark draft model checkpoints for three LFM2.5 models, enabling speculative decoding that speeds up inference by up to 3.18x on H100 GPUs and 2.87x on-device, with day-one support for llama.cpp and SGLang.
Amazon Bedrock AgentCore now lets teams write agent governance rules in plain English and automatically convert them into Dogwood, an open-source policy language, including support for time-based and cumulative constraints. The update extends Policy Authoring beyond simple parameter restrictions to rate limiting, sequential tool-call ordering, and Amazon Bedrock Guardrails integration.
Slack has launched Slack Code, a feature that gives AI coding agents dedicated project channels where developers and non-technical staff can watch, review, and redirect their work in real time. Anthropic, GitHub, OpenAI, Cognition, and Vercel are building integrations, bringing Claude, Copilot, ChatGPT, and Devin into Slack's group chat environment.
Mistral has released Agentic Search, a retrieval layer that lets AI models navigate, read, and verify information across complex documents instead of relying on single-pass chunk retrieval. The company claims accuracy improvements from 26.7% to 86% on FinanceBench and up to 39.6% lower p90 latency.
A new industry analysis argues the performance gap between Chinese and Western AI models has narrowed to single-digit differences on broad benchmarks. Moonshot's Kimi K3 and Zhipu's GLM-5.3 now trail OpenAI and Anthropic's top models by only a few points on the Artificial Analysis Intelligence Index, with a clear Western edge remaining only in abstract reasoning, output reliability, and offensive cybersecurity capability.
Tencent has released Hy-MT2-30B-A3B, a mixture-of-experts translation model with 30B total parameters and 3B active parameters, supporting 33 language pairs and five Chinese dialect and minority-language pairs. The model is available through Tencent Cloud at $0.074 per 1M input tokens and $0.295 per 1M output tokens.
Adobe has moved Firefly's Generate Music, Generate Speech, and Generate Sound Effects tools out of beta and into general availability on desktop and mobile web. The tools produce commercially licensed audio without requiring a separate subscription, and Firefly has also added Gemini Omni Flash, Runway Aleph 2.0, and Kling 2.0 to its third-party model roster.
Robotics startup Generalist AI has released GEN-1.5, a model that loads a short video demonstration into its context window and performs the task without additional training. The company reports a 59 percent success rate zero-shot and 83 percent after light fine-tuning, though all results are self-reported.
Black Forest Labs has released FLUX Upscale, a dedicated tool that takes existing video and regenerates it at higher resolution, up to native 4K. The tool is now available as its own FLUX endpoint.
Meta released a new Mac app for Meta AI with system-wide dictation and the ability to answer questions based on what's visible on screen, using its Muse Spark model. The launch also bundles business-focused features letting merchants connect Instagram, Facebook, ad accounts, and Google Workspace to the assistant.
OpenAI has built a system called Private Safety Processing that detects misuse patterns across multiple interactions without storing customer inputs or outputs. The company says it only receives narrow safety signals—type and severity of activity—while data stays encrypted on customer infrastructure.
Z.ai CEO Jie Tang says raw parameter counts no longer predict model quality, pointing to GLM 5.3's benchmark gains that came entirely from reinforcement learning on synthetic long-horizon environments rather than scaling weights. The claim lands alongside a broader reshuffling of agent and legal benchmark leaderboards.
OpenAI is previewing Private Safety Processing to select customers, an automated system that monitors for misuse across multiple sessions without retaining any customer data. The move directly contrasts with Anthropic's July policy allowing 30-day data retention for 'covered models' like Fable.
Amazon Bedrock AgentCore's Web Search connector now supports per-request domain filtering and published-date windows, enforced server-side. The update, shipped as connector version 1.2.0, also expands availability to AWS regions in Dublin and Tokyo.
Google announced new study features across Search and Gemini, including AI-generated interactive visuals, 3D simulations, customized practice quizzes, and a dedicated learning hub. The rollout intensifies competition with OpenAI and education startups like Knowt and Gauth for student users.
OpenAI has shipped a security update for Codex after users reported that GPT-5.6 Sol was autonomously deleting real files instead of temporary ones. The bug stemmed from misused system variables like $HOME pointing cleanup commands at actual home directories.
Meta has released a native Mac desktop app for Meta AI, built with AppKit and SwiftUI rather than Electron, weighing just 16MB. The 1.0 beta adds Mac-specific features including a global composer shortcut, system-wide dictation, and window-attachment for screen context.
OpenAI has reaffirmed its Zero Data Retention (ZDR) policy for eligible API customers using frontier models and previewed a new feature called Private Safety Processing, which the company claims allows safety monitoring without retaining customer data.
OpenAI has reaffirmed its Zero Data Retention (ZDR) policy for eligible API customers using frontier models and previewed a new capability called Private Safety Processing. The company says the new approach aims to preserve safety monitoring capabilities without requiring data storage.
Meta is rolling out a standalone Mac desktop app for Meta AI, adding screen-sharing capabilities and business-focused features like ad analytics and Google Workspace integration. The app is free, though advanced features may require a Meta One subscription.