Artificial Intelligence News Live: North Carolina adopts artificial intelligence guidebook for K-12 education

North Carolina has adopted an artificial intelligence guidebook for its K-12 education system, becoming the fourth U.S. state to implement such resources designed to be helpful for both teachers and students, CNN says in a report. This initiative follows earlier adoptions by states like South Carolina, West Virginia, and Maine of similar K-12 AI guidelines, such as those developed by the Council of Chief State School Officers (CCSSO) and the Council of the Great City Schools (CGCS) in late 2023. The North Carolina guidebook is intended to offer recommendations for educators on responsibly integrating AI, covering definitions, ethical considerations, data privacy, and curriculum integration to foster AI literacy and support classroom application.

Nvidia Chief Executive Officer Jensen Huang has described the UK’s artificial intelligence talent, including its researchers, universities, and startups, as the “envy of the world,” according to a report in Bloomberg. Huang concurrently emphasized that the nation requires greater investment in infrastructure, specifically high-performance computing capabilities, to fully harness this potential and support the continued development of its AI ecosystem.

Chinese regulators are reportedly seeking to slow the deployment of self-driving features in cars as Beijing faces pressure to establish world-leading safety and liability standards for the rapidly advancing technology, Financial Times says in a report. This move reflects a cautious approach to integrating complex AI, encompassing both advanced driver-assistance systems (ADAS) and higher levels of automation, into consumer vehicles. The aim is to ensure robust safety validation and clear legal frameworks for liability before widespread adoption, amidst ongoing global debates and increased scrutiny over the safety of such systems.

OpenAI claims it has recently achieved $10 billion in annual recurring revenue (ARR), up from about $5.5 billion in the previous year, TechCrunch says in a report. This growth is attributed to the scaling of its paid offerings such as ChatGPT Plus subscriptions and enterprise solutions, its deepened partnership with Microsoft including Azure integrations and Copilot products, and the introduction of new AI models and features like GPT-4o. The reported $10 billion figure represents projected revenue over a 12-month period based on current subscriptions and contracts, not actual profit, and is a claim made by OpenAI itself rather than an independently audited financial result.

Apple researchers have developed an AI image generation system named STARFlow, which is described as challenging existing diffusion models such as DALL-E and Midjourney by offering competitive performance, VentureBeat said in a report. This development signifies Apple's research activity in generative AI, with STARFlow currently being a research project rather than an announced consumer product. The report indicates STARFlow offers 'competitive performance', although specific metrics, the basis of comparison with other models, and the precise technical nature of its challenge to diffusion models are not yet fully detailed.

At its WWDC 2025 event, Apple released an updated version of Xcode, its integrated development environment, which now features OpenAI's ChatGPT to assist with coding tasks, documentation generation, and more, according to a report in TechCrunch. The company also announced that developers will be able to use API keys to connect AI models from other providers directly into Xcode, enabling AI-powered programming suggestions and expanding the artificial intelligence capabilities available within Apple's development suite.

Swedish artificial intelligence startup Lovable, which aims to streamline software development, is in talks for a significant funding round that could result in a $1.5 billion valuation, Bloomberg says in a report. This potential valuation, if achieved, would establish Lovable, known for its 'Vibe Coding' approach, as one of Europe's largest AI newcomers.

Canadian software networking startup Tailscale Inc. is developing a long-range strategy for an initial public offering, following a period where the artificial intelligence boom significantly increased its revenue and doubled its valuation to $1.5 billion, Bloomberg says in a report. The company's secure mesh networking technology is reportedly beneficial for AI development teams by simplifying secure access to distributed computing resources essential for training models and managing data. This surge in demand, driven by AI adoption, has strengthened Tailscale's market position as it considers a public offering.

Tools for Humanity, a company co-founded by OpenAI’s Sam Altman, is introducing its Worldcoin iris-scanning Orb devices in the United Kingdom as part of a global expansion of its identity verification project, Bloomberg said in a report. The Worldcoin system uses iris recognition to establish 'proof of personhood,' which is intended to verify users as unique humans, potentially granting them access to its network and the WLD cryptocurrency. This expansion into the UK means Worldcoin's biometric data collection and processing practices will likely face scrutiny from regulatory bodies concerning compliance with data protection laws, such as UK GDPR, echoing data privacy concerns that have led to regulatory actions in other countries.

Apple unveiled watchOS 26 at WWDC 2025, introducing several new features powered by its 'Apple Intelligence' system, The Verge says in a report. A key AI integration is the 'Workout Buddy,' which uses fitness data to provide personalized insights, advice, and motivational pep talks during exercise sessions. 'Apple Intelligence' also enhances the Messages app on Apple Watch Series 9 or later (when paired with an AI-enabled iPhone) with instant, automatic translation for texts and smarter, context-aware Smart Replies and suggestions. Furthermore, the Smart Stack will exhibit AI-driven contextual behavior, offering timely app suggestions based on user patterns and location.

While many corporations have actively adopted Artificial Intelligence aiming for productivity improvements, recent transaction data from fintech company Ramp indicates that the pace of this adoption, specifically corporate spending on external AI tools and services, might be leveling off, TechCrunch says in a report. This analysis suggests a potential slowing in the rate of increase of AI-related corporate expenditure tracked through Ramp's platform, rather than a complete halt or decline in all AI adoption efforts across businesses.

Apple is launching a new initiative named the Foundation Models framework, which the company states will allow developers to access and integrate its artificial intelligence models directly on devices for offline operation, TechCrunch says in a report. This framework is designed to provide developers with tools for on-device AI processing.

Google Gemini has introduced a new capability allowing users to schedule specific actions to be performed on a recurring basis, such as every hour, day, or week, ZDNet says in a report. This feature enables Gemini to autonomously carry out these tasks at the designated times, enhancing its functionality as an AI assistant. The development brings Gemini's task automation features more in line with competitors like OpenAI's ChatGPT, which already offers similar options for recurring actions.

Apple has confirmed a list of new features set to arrive with iOS 26, macOS 26, and other operating system updates later this year, with a notable "big AI one" among them, ZDNet said in a report. This confirmation comes ahead of the company's Worldwide Developer Conference (WWDC) 2025, signaling Apple's increased focus on integrating artificial intelligence capabilities across its ecosystem.

Shares of companies providing artificial intelligence infrastructure are experiencing a significant rebound, driven by renewed, substantial spending from major technology firms, which has bolstered investor confidence after a market downturn earlier this year, Bloomberg says in a report. This increased capital expenditure by tech giants like Microsoft, Google, Amazon, and Meta is primarily for acquiring essential hardware, such as high-performance GPUs and specialized AI chips, and for expanding data center capabilities. The renewed investment underscores the ongoing high demand for compute power necessary for developing and deploying advanced AI models.

Meta is reportedly planning a significant multi-billion dollar investment into a leading startup in the AI data market, believed to be Scale AI, Bloomberg said in a report. This potential investment follows earlier calls for strengthening US AI capabilities; Scale AI CEO Alexandr Wang testified before US policymakers in April 2024, advocating for a "national AI data reserve," sufficient power for data centers, and consistent national AI regulations to maintain US competitiveness, particularly in light of advancements such as a model from Chinese developer DeepSeek, which reportedly rivaled top US models earlier in 2024.

BP is placing artificial intelligence at the core of its strategy to improve operational performance, particularly in optimizing the complex process of oil drilling, according to a report in Bloomberg. The company is leveraging AI as a decision-support tool to help engineers determine the most effective routes for drilling wells by analyzing numerous potential paths based on geological data and operational constraints. This application aims to enhance efficiency in reaching underground targets and improve resource extraction by identifying optimal solutions.

OpenAI has released an update for ChatGPT's voice mode, significantly improving the naturalness of its synthetic speech and offering new voice options, TechCrunch says in a report. This enhancement to the existing feature aims to provide users with a more engaging and less robotic conversational experience when interacting with the AI. While the article description does not specify the underlying model, this improvement is widely understood to be powered by capabilities from OpenAI's GPT-4o model, which was announced on May 13, 2024, with advanced, more natural voice functionalities.

Getty Images is set to face artificial intelligence company Stability AI in a London courtroom, with opening arguments scheduled for Monday in what is being called the first major copyright trial for the generative AI industry, AP said in a report. The trial at the British High Court, which could last for three weeks, concerns Stability AI's widely used AI image-making tool, which was released in August 2022; Stability AI is based in London.

Pace University has adopted a system using artificial intelligence and scannable QR codes to read out the names of graduates at its commencement events, according to a report in Fox Business. The technology likely involves AI-powered text-to-speech to pre-generate audio files of names, based on pronunciation data provided by students, with QR codes on graduates' cards triggering the playback of the correct audio file when scanned during the ceremony, aiming to improve accuracy and personalization.

An AI system designed to enhance classroom interaction has been detailed, IEEE Spectrum says in a report. This system likely aims to improve educational engagement by providing tools that support or analyze classroom dynamics, such as identifying student participation patterns or offering prompts for discussion, thereby enabling better human interaction rather than directly replicating it. The development of such AI tools for education is an active area, though their true effectiveness is dependent on specific design, implementation context, teacher preparedness, and ensuring equitable student access, alongside addressing ethical considerations like data privacy.

OpenAI intends to appeal a specific procedural copyright-related ruling within the lawsuit filed against it by The New York Times in December 2023, which alleges infringement through the use of journalistic content for training AI models, Fox Business said in a report. Concurrently, OpenAI CEO Sam Altman has reportedly called for the creation of an 'AI privilege,' a proposed concept suggesting legal protections for certain AI development data or processes, though this is not an established legal principle.

Innatera has introduced a neuromorphic chip for artificial intelligence applications in edge devices and smarter sensors, which offers lower latency and significantly reduced power consumption, IEEE Spectrum said in a report. This chip is part of the growing field of neuromorphic computing, which mimics brain-like processing, and aims to meet the increasing demand for energy-efficient AI accelerators capable of real-time processing at the edge. By potentially using spiking neural networks, the chip is designed to enable more complex AI tasks to be performed locally on devices, enhancing responsiveness and reducing the need for cloud connectivity.

Meta is reportedly in discussions for a multi-billion dollar investment in Scale AI, an artificial intelligence data labeling and model evaluation company, TechCrunch said in a report. The potential deal value could reportedly exceed $10 billion, which would mark the Facebook parent company's largest external AI investment and a significant funding event for Scale AI. This reported interest aligns with Meta's ongoing substantial investments in AI infrastructure and development.

Research led by The Hong Kong Polytechnic University (PolyU) suggests that sensory and motor inputs significantly aid large language models (LLMs) in representing complex concepts, AP said in a report. This finding contributes to the ongoing debate in cognitive science regarding embodied cognition—the theory that physical, sensory experience is crucial for concept formation—versus the observation that LLMs can develop sophisticated world representations from language data alone. The study implies that while LLMs demonstrate an ability to learn from text, integrating or simulating sensory and motor experiences could lead to a deeper and more robust understanding in AI.

Apple is reportedly facing an uphill battle in its attempt to upgrade its artificial intelligence (AI) voice assistant, Siri, PYMNTS said in a report. This reported struggle relates to an effort that was previously linked, as of last year, to a strategy aimed at increasing hardware sales.

The British Film Institute (BFI) has released a report stating that artificial intelligence companies are utilizing approximately 130,000 film and TV scripts to train their AI models, characterizing this as 'plundering' copyrighted material and posing a 'direct threat' to the UK's screen industries, The Guardian says in a report. The BFI's analysis also highlights concerns that AI-driven automation could lead to the elimination of entry-level positions across film, television, video games, and visual special effects, thereby jeopardizing the influx of new talent into these sectors.

Internet safety campaigners in the UK have called on Ofcom, the communications regulator, to restrict the use of artificial intelligence in vital risk assessments after a report emerged last month suggesting Mark Zuckerberg's Meta intends to automate as much as 90% of such checks for its platforms, including Facebook, Instagram, and WhatsApp, according to a report in The Guardian. Ofcom has acknowledged the concerns raised in the campaigners' letter and stated it is "considering the concerns."

Major Chinese technology companies have temporarily disabled certain artificial intelligence functions to prevent cheating during the country's highly competitive gaokao university entrance exams, which began on Saturday, June 7, 2025, The Guardian says in a report. More than 13.3 million students are participating in the four-day tests for a limited number of university placements. While the specific companies and the exact AI functions affected were not detailed, the suspension likely targets generative AI capabilities, such as text generation or question-answering, that could be exploited for academic dishonesty. This action aligns with China's broader regulatory stance on generative AI, which has included rules since 2023 focusing on content control, data security, and preventing misuse, and also reflects global concerns among educators about AI's impact on academic integrity.

Mark Read is stepping down as Chief Executive Officer of WPP, the UK’s largest advertising group, as the company and the broader advertising sector grapple with ongoing challenges presented by the rapid advancements in artificial intelligence, the Financial Times said in a report. These challenges include the widespread adoption of generative AI in marketing, heightened competition from technology firms and consultancies leveraging AI, and clients increasingly developing their own AI capabilities, all of which pressure traditional agencies to adapt their models and demonstrate value in an AI-driven market.

The U.S. Food and Drug Administration has approved an artificial intelligence-powered technology tool intended for applications in cancer care, according to a report in Fox News. This approval signifies a continued trend of integrating AI into healthcare, particularly within oncology for tasks such as diagnostic assistance, image analysis, or treatment planning support, rather than serving as a direct therapeutic agent. The FDA's clearance is for a specific intended use, underscoring the growing role of AI in potentially enhancing early detection and personalizing patient care.

OpenAI aims to embed its Artificial Intelligence technology, specifically ChatGPT, into multiple aspects of university education, with an initial large-scale implementation planned for 460,000 students at California State University, The New York Times said in a report. This initiative reflects a growing trend of collaborations between AI companies and higher education institutions seeking to integrate AI into learning platforms, research, and administration. The plan signifies a broader ambition for OpenAI to become a solution provider in education, moving beyond its initial association with chatbot-related academic integrity concerns, and involves a more structured integration than simply directing students to the public ChatGPT website.

A senior UK judge warned on Friday that lawyers could face prosecution for submitting court material "hallucinated" by artificial intelligence tools, the New York Times said in a report. This statement underscores the risk legal professionals face if they fail to verify AI-generated content, which can include fabricated legal precedents or facts. In the legal context, "prosecuted" likely signifies professional disciplinary measures, contempt of court proceedings, or other legal consequences for misleading the court, rather than necessarily criminal charges. This UK development echoes situations in the US where lawyers have been sanctioned for citing AI-generated fake cases, and legal professional bodies are increasingly issuing ethical guidelines on the responsible use of AI.

A new US government report cautions that advanced Artificial Intelligence systems could pose an "extinction-level threat" to humans and states that the US must intervene, CNN says in a report. Commenting on the warning, James Stavridis acknowledged the concern but drew parallels to past transformative inventions like the printing press, electricity, and the internet, which also initially faced fears regarding potential misuse.