AI News — Saturday, July 11, 2026
OpenAI announces GPT-5.6, a new frontier intelligence model designed to offer enhanced capabilities and scalability for various applications.
Apple has filed a lawsuit against OpenAI, accusing the AI research company of stealing trade secrets, marking a significant legal escalation in the AI industry.
Microsoft 365 Copilot has integrated OpenAI's newly released GPT-5.6 as its preferred model, indicating a major upgrade in its AI-powered productivity features.
Meta has removed a controversial AI feature from Instagram following user backlash, highlighting the challenges of integrating AI into social platforms.
Anthropic introduces Cowork, a new Claude Desktop agent designed to interact directly with user files without requiring any coding, aiming to enhance productivity for a broader audience.
Salesforce has launched a new Slackbot AI agent, intensifying its competition with Microsoft and Google in the rapidly expanding workplace AI market.
Clem Delangue, CEO of Hugging Face, emphasizes the increasing importance of open-source AI in the current technological landscape, advocating for collaborative development.
Researchers introduce Video-Oasis, a new framework that re-evaluates and aims to improve the assessment of video understanding models.
This research explores methods to mitigate object-driven shortcuts in AI models for zero-shot compositional action recognition, improving their ability to understand complex actions.
A developer shares insights on building a robust error model to handle the diverse failure modes of various AI providers, moving beyond simple status code checks.
A new paper introduces a system capable of generating infinite virtual worlds with diverse and versatile interaction possibilities, pushing boundaries in generative AI.
UniClawBench is presented as a universal benchmark designed to comprehensively evaluate the performance of proactive AI agents in real-world task scenarios.
This research introduces a benchmark for evaluating AI's ability to reason about scientific lineage and generate new ideas based on historical scientific development.
This article provides guidance on how to optimize websites to be effectively interpreted and utilized by AI agents, enhancing their interaction capabilities.
LongE2V introduces a novel approach using video diffusion models for long-horizon event-based video reconstruction, prediction, and frame interpolation.