Edition: July 13, 2026

Nous Research, the startup behind the open-source Hermes agent, is in talks for a new funding round at a $1.5 billion valuation, aiming to raise at least $75 million. The funding, led by Robot Ventures, will help expand Hermes' products and business model. Hermes has gained popularity for its automated task capabilities and AI agent features.

Apple's revamped Siri AI is now available to the public as part of the iOS 27 beta, offering a chatbot-style app and deeper integration into the iPhone user experience. The new Siri AI aims to be more helpful for navigating the phone and performing tasks beyond voice-based interactions. With its expanded capabilities, Siri now feels like a fundamental part of the iPhone's infrastructure.

Microsoft CEO Satya Nadella has issued a warning to companies using AI, stating that they are paying twice, once with money and again with valuable business data, which can be used by AI model makers to become competitors. Nadella argues that companies should retain ownership of their data and build proprietary learning environments to protect themselves. He suggests building orchestration layers to easily switch between AI models from different providers.

Apple has filed a trade secrets lawsuit against OpenAI, alleging a coordinated effort to extract confidential information from current and former Apple employees, with claims of misconduct that Apple says is 'rotten to its core' and part of OpenAI's culture. The lawsuit details extraordinary allegations, including employees accessing Apple's network storage and bringing Apple parts to OpenAI interviews. Apple suggests that the alleged misconduct is just the 'tip of the iceberg' and expects to uncover more during the discovery process.

Anthropic, a leading AI company, has made a groundbreaking discovery in understanding how large language models work, revealing a hidden space where models 'puzzle over concepts' and influence their outputs. The finding, dubbed the J-space, contains words that don't appear in the model's output but affect its decision-making process. This discovery sheds new light on the complex mechanisms of AI models.

Apple has filed a lawsuit against OpenAI, alleging the company stole confidential documents, spied on hardware prototypes, and tricked a partner into sharing proprietary design techniques. The lawsuit centers around the actions of three individuals, including a former Apple vice president. The claims include unusual requests, such as asking Apple employees to bring unreleased product samples to job interviews.

Comma AI founder George Hotz has sparked debate by suggesting that AI should be designed to prioritize individual freedom, even if it means allowing users to plan harmful acts. Hotz argues that truly aligned AI should be able to assist users without judgment, comparing it to a gun that doesn't question its user's intentions. However, critics argue that such an approach neglects the need for accountability and consideration of collective well-being.

Defenders are now utilizing prompt injections to counter AI hacking agents by embedding commands that shut down attacks, a tactic previously used by attackers to exploit large language models. Researchers from Tracebit found that placing prompt injections alongside sensitive data can stop AI hacking agents. This approach directs the attacking LLM to perform forbidden actions, causing it to shut down.

India's Atal Tinkering Labs (ATL) is launching ATL Saathi, a Gemini-powered web application that provides educators with a 24/7 planning and training assistant, leveraging AI to drive innovation and learning. The initiative aims to empower over 1.1 crore students across India with access to new technology and mentorship. The launch is part of the government's effort to promote a culture of innovation and entrepreneurship in the country.

World models, a new category of AI, aim to simulate the physical world, offering a promising alternative to large language models. This emerging field has garnered significant research and funding, with more developments expected in the coming months. World models could potentially lay the groundwork for AI systems that interact with the physical world.

Cloudflare is introducing new rules requiring AI agent crawlers to obtain permission to access certain web pages, particularly those with ads, starting September 15. The change affects various types of crawlers, including search, agent, and training bots. Publishers and AI operators must navigate these new rules to maintain access to relevant content.

Business leaders are reevaluating their use of AI due to rising costs, with 29% of senior executives struggling to understand the operating costs of AI deployments and nearly half considering rephasing their deployments when costs outweigh expected value. The shift to usage-based billing has led to 'sticker shock' for many companies. As a result, companies are exploring lower-cost models and alternative solutions to mitigate the financial impact.

The first public beta of iOS 27 has been released, featuring a 'Snow Leopard update' with a focus on fixing issues and improving speed, and introducing new features like in-line replies and end-to-end encryption in the Messages app. Siri AI is also being highlighted as a key aspect of the new operating system. The update promises faster app launches and improved performance across the board.

Sam Altman and Elon Musk exchanged comments on social media regarding the feasibility of space data centers, with Altman expressing skepticism about the near-term viability of the business. Experts agree that cheaper rockets and mass production of high-powered satellites are needed for space data centers to become a reality. SpaceX's plans for orbital data centers are a key driver of the company's valuation, but significant technical and economic hurdles remain.

Anthropic has started localizing pricing for its AI service Claude in India, its largest market outside the US, with rupee-denominated subscriptions now available. The move aims to reduce friction for Indian users who previously had to pay in dollars. Prices in India include local taxes and vary slightly from US prices.

Waze has introduced new AI-powered features, including conversational road updates and personalized navigation, to enhance user experience. The app now suggests routes based on a user's trip history and traffic patterns, and also features a new Motorcycle mode for two-wheeler-specific routes. Additionally, users can report road updates conversationally and use a 'less chatty' mode for fewer interruptions.

A European fertility group is calling for international limits on the number of children a single sperm donor can contribute to, citing concerns over donor-conceived individuals having numerous unknown siblings. Meanwhile, researchers are developing 'world models' to help AI understand and operate within physical spaces. Apple has also sued OpenAI for allegedly stealing trade secrets to develop consumer hardware.

Waze is introducing new AI-powered features, including conversational reporting and destination search, powered by Google's Gemini AI assistant. These updates aim to enhance the driving experience by allowing users to personalize their trips. The features will utilize conversational voice commands for reporting incidents and finding destinations.