NerdNews

Apple is testing a standalone app for its overhauled Siri, which will debut at WWDC 2026. The new version of Siri will be a systemwide AI agent with deep integration across applications, allowing it to control other apps and features of iOS and macOS, and complete user-requested tasks using personal data. The app will provide a dedicated place to find previous interactions, search through past chats, and start new conversations.

Anthropic's latest update to Claude Code introduces 'auto mode', which allows the AI to decide which actions are safe to take on its own, with some limits. This feature uses AI safeguards to review each action before it runs, checking for risky behavior and prompt injection. Auto mode will roll out to Enterprise and API users in the coming days.

OpenAI is pivoting away from its 'Instant Checkout' feature in ChatGPT due to low adoption, instead focusing on product discovery and providing more detailed product information to users.

Anthropic is suing the Department of Defense over a supply chain risk designation, the Trump administration is posting war memes on social media, and AI may potentially replace venture capitalist jobs. The lawsuit claims the designation is damaging Anthropic's business and infringing on its free speech rights. Meanwhile, the Trump administration's use of war memes has been criticized for gamifying war and being insensitive to the casualties. Additionally, a new platform called ADIN is using AI agents to analyze startup pitch decks and make investment recommendations, potentially disrupting the venture capital industry.

Meta has been found liable in a New Mexico court case for misleading consumers about the safety of its platforms and endangering children. The company has been ordered to pay $375 million in civil penalties. This verdict may have significant implications for Meta and other social media companies, with thousands of similar cases pending.

The rise of smart devices and biometric surveillance has made Americans more vulnerable to police searches. The use of face recognition, DNA matching, and other technologies has raised concerns about privacy and autonomy. As these technologies become more prevalent, it is essential to consider the implications for individual rights and freedoms.

A New Mexico jury ordered Meta to pay $375 million in civil penalties for misleading consumers about the safety of its platforms and endangering children. The verdict is the first of its kind against Meta and may have implications for the company's handling of child safety on its platforms.

Databricks has acquired two startups, Antimatter and SiftD.ai, to support its new AI security product, Lakewatch. Lakewatch uses AI agents to detect and investigate threats, and the acquisitions will help Databricks expand its security capabilities. The terms of the deals were not disclosed.

A new hacking group, TeamPCP, is spreading self-propagating malware that poisons open-source software and wipes machines based in Iran. The malware, named CanisterWorm, targets CI/CD pipelines and uses a novel wiper that only activates on machines in the Iranian timezone or configured for use in that country.

Uber's uSpec uses AI agents and Figma Console Model Context Protocol to automate component design specifications, reducing documentation time from weeks to minutes.

HP IQ is a local AI and collaboration application that enables features like chat, file sharing, and meeting summarization. It uses OpenAI's gpt-oss-20b model and is initially available on HP's 2026 EliteBook and ProBook models. The feature raises potential privacy concerns, but HP recommends users follow best practices and ask for permission before recording meetings.

Amazon has acquired Fauna Robotics, a startup developing kid-size humanoid robots. The acquisition brings Fauna's employees and technology to Amazon, with potential applications in consumer robotics and home automation.

Agile Robots has partnered with Google DeepMind to develop robots using Gemini Robotics foundation models. The partnership aims to test, fine-tune, and deploy robots in industrial use cases, including electronics manufacturing and logistics. This collaboration is the latest in a string of robotics companies partnering with Google DeepMind to advance their technology.

Agoda observes that AI coding tools have increased individual developer output, but project-level velocity gains are modest because coding was never the bottleneck. The bottleneck has shifted to specification and verification, requiring human judgment. This shift implies changes in team structure, with a focus on collaborative specification and architectural alignment.

Arm, a leading chip design firm, is producing its own semiconductors, including an AI CPU called the Arm AGI CPU, which is designed to be energy-efficient and handle agentic AI tasks. The chip is being fabricated by Taiwan Semiconductor Manufacturing Corporation and has already garnered interest from major customers like Meta, OpenAI, and Cloudflare.

Apple's iOS 26.4 update introduces AI-generated playlists, concert discovery, and improved accessibility features. It also adds new emoji and enhances the Family Sharing feature, allowing each adult member to add their own payment method.

Brinc, a startup founded by a former Thiel Fellow, has launched a drone called Guardian that can replace police helicopters. The drone has a 62-minute flight time, thermal imaging cameras, and a loud speaker. It also comes with a Starlink panel for unlimited range and can be equipped with critical safety supplies.

Zoox is launching its robotaxi service in Austin and Miami, and expanding its service areas in San Francisco and Las Vegas. The company has driven nearly 2 million autonomous miles and carried over 350,000 riders. Zoox is also working to remove federal restrictions to offer paid rides and has partnered with Uber to launch commercial services in Las Vegas.

Epic Games, the company behind Fortnite, is laying off over 1,000 employees due to financial struggles. Despite the layoffs, CEO Tim Sweeney claims that AI is not the cause, instead citing a slowdown in Fortnite engagement and broader industry challenges. The company is also implementing a $500 million cost savings plan.

OpenAI is shutting down its Sora video generation app just 15 months after its launch. The move comes as the company refocuses on business and productivity applications. Sora was a photorealistic video generator that allowed users to create complex videos with ease.