News

Hachette Book Group, Cengage Learning and Elsevier, joined by author Scott Turow, have lodged a class‑action lawsuit against Google, accusing the company of illegally using millions of copyrighted works to train its Gemini artificial‑intelligence system. The complaint alleges that Google harvested the material without permission, stripped copyright management information and failed to implement safeguards, effectively allowing Gemini to generate derivative content without credit or compensation. The suit joins a growing wave of legal challenges targeting AI firms for alleged infringement of publishing and literary rights.Read more

SpaceXAI’s Grok Build AI coding assistant was found to be uploading entire user code repositories—including deleted files and hidden secrets—to Google Cloud storage. The behavior triggered a swift response from the company, which disabled the upload function after security researchers highlighted the privacy risk. Elon Musk assured users that all captured data would be erased, but experts warned that the volume of information exposed could include proprietary source code, credentials, and infrastructure details. The incident underscores ongoing concerns about data handling in AI development tools.Read more

A security researcher discovered that xAI’s Grok Build coding CLI uploaded entire tracked Git repositories—complete with commit history, API keys and other secrets—to a Google Cloud Storage bucket. The upload volume was about 27,800 times larger than the code needed for the AI task, directly contradicting the tool’s claim that no code data leaves a user’s machine. Elon Musk confirmed the breach and said the company would erase all prior Grok Build data, but independent verification of the deletion remains pending.Read more

Anthropic announced a CAD$10 million investment to support AI research at eight Canadian institutions, including the country’s leading AI hubs in Edmonton, Montréal and Toronto. The funding will back projects ranging from reinforcement learning and AI safety to mental‑health modeling, Indigenous language tools and quantum‑computing applications. Canada ranks second worldwide in per‑capita usage of Anthropic’s Claude model, a metric the company highlighted in its new Canadian country brief. Startups linked to the partner institutes will also receive at least $5,000 in API credits.Read more

Chinese open‑weight models captured 41% of Hugging Face downloads this spring, overtaking U.S. offerings. The shift reflects a growing preference among developers and enterprises for customizable, cost‑effective AI, with open models now handling nearly a third of AI requests on platforms like Vercel. Industry leaders, from Hugging Face CEO Clem Delangue to Microsoft’s Satya Nadella, warn against lock‑in to closed‑source providers and argue that broader access could balance power in the AI ecosystem.Read more

Chinese AI startup DeepSeek is preparing a $1.5 billion fundraising round that could value the company at about $71 billion, Bloomberg reports. The capital push is tied to an IPO plan slated for 2027, though the company may list as early as the end of this year. DeepSeek, founded in 2023, recently raised $7 billion at a $50 billion valuation and now commands roughly 23% of token traffic on Vercel’s enterprise AI gateway, trailing Anthropic’s 32%. Investors include Tencent and Beijing’s National Artificial Intelligence Industry Investment Fund.Read more

A coalition of major publishers, including Hachette, Cengage and Elsevier, along with author Scott Turow and the author‑rights group S.C.R.I.B.E., has filed a class‑action lawsuit in New York federal court accusing Google of illegally copying their copyrighted works to train its Gemini artificial‑intelligence model. The plaintiffs claim Google stripped copyright information to hide the training process and violated long‑standing agreements that limit Google’s use of books to searchable snippets. The case adds to a growing wave of copyright suits targeting AI firms, amid mixed rulings from California courts.Read more

Anthropic’s latest television spot, titled “There’s hope in hard questions,” has ignited a firestorm on social media. The ad’s unsettling visuals—a burning house, surveillance footage, a homeless sleeper, rows of tombstones and a mining scene—paired with voice‑over questions about AI trust, have prompted accusations of poor corporate communication. OpenAI chief Sam Altman and other tech observers mocked the piece, calling it creepy and tone‑deaf, while the company defends the campaign as an extension of its ethical‑AI branding.Read more

Google DeepMind chief Demis Hassabis urged the creation of a self‑governing regulator modeled after FINRA to vet and certify frontier artificial‑intelligence models before they hit the market. In a LinkedIn post, he outlined a framework that would give labs a 30‑day review window, shift release decisions away from ad‑hoc government panels, and fund the body through industry contributions.Read more

AI startup Nous Research, creator of the open‑source Hermes agent, has closed a new funding round led by Robot Ventures with participation from USV and other investors. The company is raising at least $75 million at a $1.5 billion post‑money valuation, according to sources familiar with the deal. The capital will fund expansion of Hermes’ product lineup and its cloud‑hosted services. Founded in 2023, Nous Research has already attracted $70 million from backers such as Paradigm and North Island Ventures.Read more

Apple filed a lawsuit on July 11, 2026, accusing OpenAI and two former Apple engineers—Tang Yew Tan and Chang Liu—of pilfering proprietary designs and confidential data to accelerate OpenAI’s hardware ambitions. The complaint says Liu left Apple with a company laptop, exploited a network flaw, downloaded unreleased product files and bragged about the breach, while Tan allegedly used Apple’s secret metal‑finishing techniques and internal codenames to aid OpenAI. OpenAI’s communications director dismissed the allegations, stating the company has no interest in Apple’s trade secrets. The case, slated for a jury trial, could reshape the fragile partnership between the two tech giants.Read more

Apple filed a 41‑page lawsuit against OpenAI, claiming the AI firm lured three former Apple engineers to help it develop its first hardware device. The complaint alleges that the ex‑employees accessed Apple’s cloud storage after leaving, downloaded confidential design files, and coached new hires on how to evade Apple’s exit procedures. Apple also says OpenAI misled a trusted supplier into believing it had permission to use Apple’s proprietary metal‑finishing process. OpenAI responded that it does not seek other companies’ trade secrets and remains focused on its own technology.Read more

Apple filed a lawsuit on Friday accusing OpenAI of using trade secrets stolen by former Apple employee Chang Liu. The complaint says Liu discovered an authentication bug after joining OpenAI, used it to download dozens of confidential hardware files, and shared the data while developing AI‑powered devices. Apple seeks injunctions to block OpenAI from using the information and demands damages. The company says the bug was quickly patched and that no other users exploited it, but it alleges a coordinated effort to shortcut Apple’s product development.Read more

Anthropic has begun offering its Claude AI service in Indian rupees, marking the first localized pricing for the market that accounts for nearly 6% of the model’s global usage. While the move brings subscription costs into the local currency, the company still relies on credit‑card and app‑store billing, leaving India’s ubiquitous UPI payment system out of the loop. Prices for Claude Pro, Claude Max and team plans now appear on the Indian version of the website, reflecting local taxes and a slight premium over U.S. rates.Read more

Tom Blomfield, the entrepreneur behind fintech firms Monzo and GoCardless, announced he is leaving his role at Y Combinator to join Anthropic as a senior member of its compute team. The move places the former digital‑bank CEO at the heart of one of AI’s biggest operational challenges—securing massive compute power for next‑generation models. Blomfield framed the job as mission‑critical, citing the potential of powerful AI to improve lives worldwide. Anthropic, already racing to deploy a million Google TPUs and hundreds of thousands of GPUs, is bolstering its talent pool with high‑profile hires as it prepares for a possible public listing.Read more

Meta unveiled a fresh AI image generator and put it through a five‑prompt showdown against OpenAI's ChatGPT and Google's Gemini Nano Banana 2. The trial covered product ads, wildlife photography, social‑media invites, comic strips and a noir‑cartoon mash‑up. While all three tools produced solid visuals, ChatGPT most consistently hit the mark, Gemini excelled at realism, and Meta delivered the most daring, humor‑laden creations. The results highlight each model’s personality and suggest that the best choice depends on whether users prioritize accuracy, photorealism or creative flair.Read more

Apple filed a lawsuit on Friday accusing OpenAI of stealing proprietary hardware trade secrets and poaching former Apple staff. The complaint cites unauthorized access to internal files, a checklist designed to help new hires evade Apple’s security protocols, and a shared manufacturing partner that allegedly used Apple’s metal‑finishing technique. Bloomberg’s Mark Gurman argues the case could slow OpenAI’s push to launch a hardware product intended to rival the iPhone, as engineers divert time to legal defenses and tighter internal controls. The dispute arrives as OpenAI prepares for an IPO and a major hardware debut.Read more

Elon Musk, fresh from a jury verdict that favored Sam Altman in their long-running legal battle, turned to X to label the OpenAI chief a “scammer.” The exchange ignited after Apple sued OpenAI for alleged trade‑secret theft, prompting Musk to launch a public feud. Altman retorted, drawing millions of views, while both companies rolled out flagship AI models—OpenAI’s GPT‑5.6 Sol and SpaceX’s Grok 4.5—underscoring a rivalry that now plays out as much in social media as in the courtroom.Read more

OpenAI announced on Thursday that its Atlas browser will be deprecated on Aug. 9, ending the standalone product that launched just months ago. The move coincides with the rollout of ChatGPT Work and a redesigned desktop app that embeds browsing, coding and productivity tools under a single interface. While headlines called the browser "dead," the company says the functionality lives on as a feature of the new app and an upgraded Chrome extension. Executives cite lessons from Atlas users and a strategic shift toward a unified super‑app rather than a separate web browser.Read more

OpenAI said its head of safety systems, Johannes Heidecke, will leave the company as it reshuffles its safety and research leadership. Heidecke, who joined in 2021, announced his exit in an internal memo. Saachi Jain will step in as interim head, while safety teams will now report to newly appointed vice president of research and safety, Mia Glaese. The changes come on the heels of the GPT-5.6 rollout and a U.S. government approval, and underscore OpenAI’s push to tightly integrate safety with frontier‑model development.Read more