Tuesday, June 30, 2026

Rockstar workers push to unionize ahead of GTA VI’s launch

Workers at Rockstar Games, the developer of Grand Theft Auto VI, have submitted a request for voluntary recognition of their union, the IWGB Game Workers Union. This move comes after the company faced backlash for firing over 30 employees last year, which was perceived as an attempt to undermine union efforts. Since its formation in 2019, the IWGB claims to represent a significant portion of the workforce across Rockstar's studios in Edinburgh, Dundee, Lincoln, Leeds, and London. The union's goals include enhancing pay transparency and addressing the notorious crunch culture within the company. A hearing regarding the previous firings is scheduled for September, just ahead of the highly anticipated GTA VI launch on November 19, priced at $79.99.

Anthropic’s Claude Science bets on workflow, not a new model, to win over scientists

Anthropic unveiled Claude Science, an AI workbench designed to streamline scientific research by integrating various databases and tools into a single environment. Unlike previous models, Claude Science does not introduce a new AI model but utilizes existing Claude models, including Claude Opus 4.8. This platform allows scientists to manage projects through a primary AI assistant that connects to over 60 databases and can create specialized sub-assistants for specific tasks. Additionally, a fact-checker AI ensures accuracy in citations and calculations, addressing concerns about AI-generated content in scientific publications.

Libby will filter out AI content, kind of

Libby, the ebook-lending app operated by OverDrive, is set to introduce AI content filters that will allow users to opt out of AI-generated materials, including AI-authored books and AI-narrated audiobooks. This initiative comes as OverDrive prepares for the potential disruption from AI-generated content in the digital publishing space. Users will need to rely on publishers to self-label their works for the filters to function effectively. DeBevoise believes that while AI poses challenges, it can also enhance access to information and improve content recommendations.

Blue Origin pivots to redesigned launchpad after explosion in push to fly by end of 2026

Blue Origin is shifting to a redesigned launchpad configuration after a New Glenn rocket explosion during a hot-fire test on May 29, 2026. CEO Dave Limp emphasized the company's commitment to returning to flight by the end of this year, stating they will not rebuild the damaged pad but will utilize existing infrastructure for a hybrid launch setup. The explosion, which occurred at Cape Canaveral Space Force Station, has prompted an active investigation into the mishap, with early indications pointing to issues in the rocket's aft section. This incident is particularly critical as Blue Origin supports NASA's Artemis program and various satellite launches for Amazon.

X now offers an MCP server to make its platform easier for AI tools to use

X has introduced a hosted Model Context Protocol (MCP) server that simplifies the integration of AI tools such as Claude, Cursor, and Grok Build with its platform. This new server allows these applications to communicate with the X API using the user's own account permissions, eliminating the need for developers to create and host their own MCP servers. While the hosted MCP does not add new capabilities, it streamlines access to existing API functionalities, positioning X as a real-time information network. This move aligns X with other companies offering MCP servers, including GitHub and Slack, but raises concerns about potential increases in automated posting and spam.

AWS puts $1 billion into new AI unit to embed engineers with customers, joining growing wave

Amazon Web Services (AWS) has announced a significant investment of $1 billion to create a new Forward Deployed Engineering (FDE) unit aimed at helping customers build and implement artificial intelligence systems. This initiative will involve embedding 'thousands' of FDEs within client organizations, with initial teams of approximately five or six engineers working directly alongside customer staff. AWS' vice president, Francessca Vasquez, emphasized that this structured approach is designed to accelerate technical transformations and deliver rapid value to stakeholders, addressing the increasing demand for speed in AI deployment.

Amazon launches new $1 billion FDE org, following OpenAI and Anthropic

Amazon Web Services (AWS) has introduced a new internal organization dedicated to forward-deployed engineers (FDE), committing $1 billion to enhance AI integration for clients. This initiative aims to embed engineers within client companies to deploy customized AI agents, ensuring rapid engagement and fostering customer self-sufficiency. AWS VP Francessca Vasquez stated that clients will not only receive new AI solutions but also gain lasting engineering skills and workflows to innovate independently. This model, originally pioneered by Palantir, allows for tailored deployments while leveraging reusable technology across different clients.

Jun 30, 2026 Announcements Claude Science, an AI workbench for scientists, is now available

Anthropic has launched Claude Science, an AI workbench designed to streamline scientific research by integrating various tools and databases into a single platform. This new app allows researchers to analyze literature, execute complex research tasks, and produce auditable artifacts, all while managing computing resources efficiently. Currently in beta for Claude Pro, Max, Team, and Enterprise users, Claude Science features over 60 pre-configured skills tailored for fields such as genomics and cheminformatics, enabling scientists to conduct their work more effectively and reproducibly.

Cheniere Energy (LNG) Benefited from Supply Constraints

Cheniere Energy, Inc. (NYSE:LNG) experienced a significant stock surge of 47% following its strong fourth-quarter earnings report, closing at $243.97 per share on June 29, 2026. The company's market capitalization reached $51.12 billion as heightened geopolitical tensions, particularly involving Iran, and structural damage to Qatari LNG facilities sidelined 20% of global liquefied natural gas capacity. This tightening of supply has led to a favorable demand environment for Cheniere, which operates LNG terminals in New Orleans and Corpus Christi, and has resulted in elevated share buybacks and increased guidance for the upcoming years.

Meituan's LongCat-2.0 shows China can train massive AI models without Nvidia

Meituan has successfully trained its LongCat-2.0 AI model, which boasts 1.6 trillion parameters, entirely on Chinese hardware, marking a significant milestone in China's AI capabilities. The training process utilized over 50,000 domestically produced AI ASICs and encompassed more than 35 trillion tokens. Despite existing US export controls since 2022, Meituan's achievement signals that China can now produce competitive AI models without reliance on Nvidia technology. While LongCat-2.0 excels in certain benchmarks, it still falls short of others, indicating a mixed performance landscape in the AI model arena.

Supreme Court strikes down limits on political parties' campaign spending, in win for GOP

The Supreme Court ruled 6-3 to eliminate limits on the amount political parties can spend in coordination with candidates, a significant victory for Republicans who claimed the restrictions violated First Amendment rights. This ruling overturns coordinated-expenditure limits that had been established for over 25 years and previously upheld in a legal challenge. The decision was opposed by the court's liberal justices, indicating a deep ideological divide on campaign finance issues.

Working Clicks Communicator unveiled in ‘first look’ as launch date gets closer [Video]

Clicks has provided a first look at the Clicks Communicator, a productivity-focused Android device featuring a physical keyboard, after months of development. While not yet finished, the device is operational, showcasing a custom Android build with a Niagara launcher that allows users to define their app lists. Notifications from favorite apps integrate seamlessly with the homescreen, enhancing usability. The Communicator also features a removable rear panel for easy access to SIM and MicroSD cards, addressing user preferences for traditional features. More updates are expected in the coming weeks as Clicks approaches its planned launch later this year.

US campaigns now run on AI at nearly every step, and Europe is drawing a harder line

AI has become a cornerstone of both Republican and Democratic campaigns in the US, with 87% of strategists employing it daily for tasks such as voter data analysis and campaign material production. A recent New York Times report highlights how campaigns are using AI to micro-target voters and vet candidates, exemplified by the Democratic group American Bridge's use of AI to analyze 250 Republican candidates. In contrast, Europe is implementing stricter regulations, mandating transparency in political advertising and requiring explicit consent for targeted campaigns.

Google releases ‘Pixel Audio Services’ for Pixel 10 on Play Store

Google has introduced the Pixel Audio Services application specifically for the Pixel 10 series, which includes the Pixel 10 Pro Fold and Pixel 10a. This system component is designed to deliver the latest updates and bug fixes for audio modules, enhancing the overall audio experience for users. The app, which is pre-installed on devices, is crucial for maintaining optimal audio performance and was last updated on June 29, 2026. However, it is not available for the older Pixel 8 or 9 series devices running Android 17.

Presentation: Trustworthy Productivity: Securing AI-Accelerated Development

Sriram Madapusi Vasudevan, a Senior Software Engineer at AWS, presented on the vulnerabilities of autonomous AI agents during his talk at QCon San Francisco. He highlighted the ReAct loop, which encompasses context management, reasoning, and tool execution, as critical points for potential failures. Vasudevan recounted a 2025 incident at Replit where a simple command to 'clean the database' led to the destruction of nine days of production data due to a misinterpretation by the AI agent. He emphasized the need for robust defense strategies, including provenance gates and mission-scoped memory, to secure AI systems against similar risks.

Redmagic’s new Android gaming tablet launches July 17 with 9-inch OLED display, 8,300 mAh battery, more

Redmagic has unveiled the Gaming Tablet 5 Pro in China, with the global launch of the Redmagic Astra 2 scheduled for July 17. This Android gaming tablet features a 9.06-inch 2.4K OLED display boasting a 185Hz refresh rate and 1,600 nits peak brightness. It is powered by a Snapdragon 8 Elite Gen 5 processor and a secondary RedCore R4 chip designed to enhance gaming performance. The device is equipped with an 8,300 mAh battery and supports fast charging, with a maximum of 80W in China, although global charging specifications are yet to be confirmed.

Mag 7 value shrinks by $2.3 trillion amid AI spending jitters — but investors are still backing chipmakers

The value of the Magnificent 7, which includes Microsoft, Nvidia, Alphabet, Apple, Meta, Tesla, and Amazon, has plummeted by $2.3 trillion this month as investors express concerns over the massive expenditures on AI infrastructure. The CNBC Magnificent 7 Index has fallen 10% in June, with Microsoft down 20% and Nvidia down 13%. Companies are collectively investing hundreds of billions in chips and data centers, but investors are now eager for returns as they await the upcoming second quarter earnings season, which begins next month.

Meta secretly tested ChatGPT, Gemini, and Character.AI with thousands of minor-perspective crisis prompts

Meta conducted secret tests on rival chatbots, including ChatGPT and Character.AI, using thousands of prompts about sensitive topics like suicide and drug use from the perspective of minors. The initiative, known internally as 'Cannes,' involved hundreds of contractors posing as minors and sending over 45,000 crisis-related prompts to these AI systems. Meta claims this practice was part of responsible safety testing and did not utilize the responses to train its own models. However, the companies being tested were unaware of this project, with Character.AI asserting that it violated their terms of service.

Former tech worker quit his 14-year career to launch a BBQ restaurant and made $2.3M in revenue in his first year

Salahodeen Abdul-Kafi, a former tech worker with a 14-year career at companies like Google and Microsoft, launched Kafi BBQ in Irving, Texas, in December 2024, generating nearly $2.3 million in revenue during its first year. He identified a gap in the Texas barbecue market for halal options, catering to a demographic that previously lacked access to such offerings. With projections of $3.6 to $4 million in revenue for 2026, Abdul-Kafi's venture reflects a growing demand for halal food, valued at $733.6 billion in the U.S. in 2025.

Law proposed to ban AI companies from selling your health data

Senator Elizabeth Warren and Representative Mary Gay Scanlon are set to introduce an updated version of the Health and Location Data Protection Act, which will prohibit AI companies from selling health data collected during chatbot interactions. This legislative move comes as AI chatbots increasingly encourage users to upload sensitive health information, raising significant privacy concerns. Notably, in January 2026, Elon Musk urged users to share their medical records with xAI's Grok, while OpenAI launched ChatGPT Health, a secure platform for health data. The proposed law aims to protect individuals from potential misuse of their health information by AI systems.

Microsoft Brings AI-Powered Vulnerability Remediation to Azure DevOps with Copilot Autofix

Microsoft has unveiled the limited public preview of Copilot Autofix for Azure DevOps, which automates vulnerability remediation for teams using Azure Repos. This new feature leverages CodeQL to analyze security vulnerabilities and uses GitHub Copilot to generate proposed fixes, creating pull requests for developers to review. By combining static analysis with AI, Microsoft aims to streamline the remediation process, addressing the significant bottleneck in secure software delivery where developers previously had to manually interpret alerts and implement fixes. Copilot Autofix emphasizes human oversight, ensuring that developers validate AI-generated recommendations before merging them into production code.

Netflix ramps up account sharing protections via new email requirement

Netflix has intensified its crackdown on password sharing by requiring that each profile on an account be linked to a unique email address. This new policy, which began rolling out approximately two weeks ago, aims to prevent unauthorized access among users who share accounts across different households. Previously, Netflix attempted to limit sharing by enforcing that all users must be located at a single address, a measure that started in 2022. While the company frames the email requirement as a way to enhance user experience with features like personalized suggestions, many users have already identified workarounds, such as using variations of Gmail addresses.

Three AirDrop vulnerabilities discovered, with Apple working on a full fix

Three vulnerabilities in AirDrop have been identified by security researchers, affecting both iPhone and Mac devices, and similar issues have been found in Android's Quick Share. These vulnerabilities allow attackers within a 10 to 30 meter range to exploit the system, causing services like AirDrop, AirPlay, Handoff, Universal Clipboard, and Continuity Camera to crash. While no data can be extracted from the devices, the services can be rendered unavailable for the duration of the attack. Apple has addressed one of the vulnerabilities, which has been assigned a CVE identifier, while work continues on the other two.

Unlocking Britain’s next era of productivity: Building a nation of AI trailblazers

AI adoption in the UK has surged to 73% in 2026, up from 34% in 2025, according to a study by Google and Public First. However, this growth is uneven, with only the top 15% of users experiencing significant career benefits such as promotions and pay raises. These 'AI Trailblazers' save nearly 8 hours each week, effectively gaining an extra workday. The challenge now lies in upskilling the remaining 85% of the workforce, many of whom are still in early-stage AI adoption, to harness these productivity gains.