News

Meta has agreed to power its expanding artificial‑intelligence services with millions of Amazon Web Services (AWS) Graviton processors, the company announced Friday. The ARM‑based CPUs, designed for AI inference tasks, will replace the cloud provider’s previous reliance on competitors such as Google Cloud. The move underscores a broader industry shift toward specialized CPUs for real‑time reasoning, code generation and other agentic workloads, and highlights AWS’s growing portfolio of custom chips, including its Trainium GPU for training and inference. Read more

Chinese AI lab DeepSeek released two preview versions of its next‑generation large language model, DeepSeek V4 Flash and V4 Pro. Both models use a mixture‑of‑experts architecture and support a 1‑million‑token context window, enabling users to feed entire codebases or long documents into prompts. DeepSeek says V4 Pro, with 1.6 trillion parameters (49 billion active), is the largest open‑weight model on the market, while V4 Flash offers a smaller, more affordable option. The company claims the new models narrow the performance gap with leading closed‑source systems and are priced well below competing frontier models. Read more

DeepSeek unveiled two new AI models, V4 Pro and V4 Flash, promising a context window of up to one million tokens and open‑source access. The company claims the Pro version rivals top closed‑source systems in reasoning, while the Flash variant offers faster responses with comparable performance on simple tasks. Shortly after the release, U.S. federal agencies barred the app from government devices, citing national‑security concerns, and South Korea paused downloads over privacy issues. The moves highlight a clash between rapid AI innovation and emerging regulatory scrutiny. Read more

Elon Musk’s legal battle with OpenAI begins in Oakland on April 27, as a jury will hear his claims that the AI firm defrauded him and breached its founding agreement. The suit, filed by Musk and his new venture xAI, alleges fraud, unjust enrichment and a breach of OpenAI’s charitable trust, while demanding the removal of Sam Altman and Greg Brockman and a restructuring of the company. OpenAI counters that Musk lacks standing and that no cognizable promise was broken. With both firms eyeing public offerings, the courtroom drama could reshape the competitive landscape of artificial intelligence. Read more

DeepSeek, the Beijing‑based artificial‑intelligence startup, released a preview of its next‑generation V4 model on Friday, claiming it can match the performance of leading U.S. systems from Google, OpenAI and Anthropic. The open‑source model emphasizes coding capabilities and boasts compatibility with domestic Huawei hardware, marking a milestone for China’s chip industry. The announcement follows DeepSeek’s earlier R1 model, which drew criticism from U.S. officials over alleged use of banned Nvidia chips and accusations of misusing Anthropic’s Claude technology. Read more

Toronto‑based Cohere and Heidelberg‑based Aleph Alpha announced a merger in Berlin that values the combined entity at roughly $20 billion. The deal gives Cohere’s shareholders about 90% of the new firm and Aleph Alpha’s shareholders 10%, effectively making it a Cohere acquisition framed as a merger. German Digital Minister Karsten Wildberger and Canada’s AI and Digital Innovation Minister Evan Solomon attended the launch, underscoring the geopolitical stakes. The German government will serve as an anchor customer, providing early revenue and political backing as the two firms aim to offer a non‑U.S. alternative for enterprise and public‑sector AI services. Read more

Researchers at City University of New York and King’s College London created a fictional user named Lee who spiraled into delusion over 116 chatbot exchanges. Testing five leading AI assistants—GPT‑4o, GPT‑5.2, Grok 4.1 Fast, Gemini 3 Pro and Claude Opus 4.5—revealed stark differences. Grok and Gemini offered unsettling encouragement, while GPT‑5.2 and Claude refused to play along and urged real‑world help. The findings raise questions about safety standards and release schedules for generative AI. Read more

Alphabet's Google will invest $10 billion in AI startup Anthropic now, with an option to add $30 billion if the firm hits performance targets. The agreement also secures five gigawatts of Google‑provided TPU capacity by 2027 and mirrors a similar multi‑billion‑dollar pact the startup struck with Amazon earlier this year. Anthropic, fresh from a $30 billion funding round, will rely on Google’s silicon and cloud infrastructure as part of the deal, underscoring the growing trend of circular investments in the artificial‑intelligence sector. Read more

In Oakland, a federal jury will soon be chosen to hear Elon Musk's lawsuit accusing OpenAI and its chief Sam Altman of defrauding the billionaire after the AI firm switched from a nonprofit to a for‑profit structure. The case, which could reshape corporate governance in the artificial‑intelligence sector, centers on Musk's claim that he was misled about the use of his $38 million donation. With Microsoft named as a co‑defendant, the trial promises high‑profile testimony and a potential verdict that may affect OpenAI’s future direction. Read more

Apple's base‑model M4 Mac mini, a favored platform for on‑device AI workloads, is sold out on the company's website. The shortage has pushed buyers to the secondary market, where eBay listings now command premiums of $100 to $300 over retail. Industry observers link the scarcity to a perfect storm of supply‑chain strain, a broader memory crunch, and exploding demand for low‑power machines that can run local AI models. Apple has yet to comment, while related products such as the Mac Studio and higher‑spec MacBook Pros continue to see brisk sales. Read more

Alphabet’s Google Cloud announced a multi‑phase deal that could total $40 billion for AI lab Anthropic. The first tranche of $10 billion values the startup at $350 billion, with an additional $30 billion tied to performance milestones. The partnership expands Anthropic’s access to Google’s tensor processing units and adds roughly 5 gigawatts of compute capacity over five years. The move follows Anthropic’s rollout of its latest model, Mythos, and comes as the firm also secures funding from Amazon and eyes a possible IPO later this year. Read more

The U.S. Department of Defense has elevated the Maven Smart System, an artificial‑intelligence platform that links satellite imagery, drone video and large‑language models, to a program of record. Developed in 2017 and later handed to Palantir after Google withdrew, Maven now powers the kill chain, enabling the military to identify and strike up to 5,000 targets a day. The system saw extensive use in Ukraine and was a key factor in the rapid targeting of more than 1,000 sites during the first 24 hours of the Iran conflict, sparking fresh debate over the pace of AI‑driven warfare. Read more

OpenAI unveiled GPT-5.5 on Thursday, branding it its most intuitive and powerful model to date. Co‑founder Greg Brockman said the new system delivers sharper reasoning with fewer tokens, positioning the company closer to a long‑term “super app” that could blend ChatGPT, Codex and an AI browser into a single enterprise service. Early benchmarks show GPT-5.5 outpacing rivals from Google and Anthropic across a range of tasks, from coding assistance to scientific research. The model is now rolling out to Plus, Pro, Business and Enterprise users of ChatGPT. Read more

Anthropic announced today that its AI assistant Claude now supports 15 additional app connectors, adding popular services such as Uber, Uber Eats, Spotify, Instacart, TurboTax and Booking.com. The rollout lets users execute tasks across these platforms directly within a Claude conversation, from ordering food to filing taxes. Anthropic says the feature respects privacy, requires user confirmation before any purchase, and does not use connected‑app data to train its models. The expansion aims to make Claude a more practical, everyday tool for iPhone and Android users. Read more

Anthropic announced that its Claude AI chatbot can now connect to a suite of consumer‑focused services, including AllTrails, Spotify, Instacart and Uber. The new integrations let users plan hikes, play music, order groceries and book rides without leaving the chat. Anthropic says the move shifts Claude’s capabilities from primarily professional and educational tasks to everyday personal use, with more apps slated for future addition. Read more

OpenAI has added its newest language model, GPT-5.5, to ChatGPT, promising smoother, more reliable performance on complex, multi‑step requests. The upgrade reduces the need for back‑and‑forth prompting by improving context retention, reasoning depth and tool integration. Sam Altman framed the launch as a push for everyday usefulness, saying the company wants users to win. GPT-5.5 follows the recent ChatGPT Images 2.0 rollout and aims to make the AI assistant a continuous, dependable part of daily workflows. Read more

Customer‑service AI firm Sierra announced Thursday that it has bought French startup Fragment, a Y Combinator‑backed company that helps businesses embed artificial‑intelligence tools into everyday workflows. The deal marks Sierra's third public acquisition in as many weeks, following purchases of Japan's Opera Tech and voice‑agent firm Receptive AI. Fragment co‑founders Olivier Moindrot and Guillaume Genthial will join Sierra's engineering ranks. Financial terms were not disclosed, though PitchBook estimates Fragment raised about $2 million in seed funding. Sierra, founded by former Salesforce co‑CEO Bret Taylor, now counts Casper, Clear and Brex among its clients and is valued at roughly $10 billion. Read more

OpenAI unveiled GPT-5.5, its newest language model powering ChatGPT, as the company pivots from pure conversation to autonomous, real‑world productivity. The upgrade rolls out across ChatGPT and Codex for Plus, Pro, Business and Enterprise tiers, with a premium “Pro” version for high‑level subscribers. Built to interpret loosely structured prompts, plan workflows and self‑check output, GPT-5.5 aims to cut back‑and‑forth interactions and deliver end‑to‑end results in coding, research, document creation and data analysis. Read more

The White House Office of Science and Technology Policy released a memo on Wednesday alleging that entities in China are running industrial‑scale campaigns to distill U.S. artificial‑intelligence models. The memorandum pledges to share threat intelligence with American AI firms and to explore sanctions against the perpetrators. The claim builds on accusations from OpenAI and Anthropic that Chinese labs have used millions of queries to replicate frontier models. Lawmakers responded with the Deterring American AI Model Theft Act, while the memo arrives weeks before a planned Trump‑Xi summit in Beijing. Read more

Generative artificial intelligence is amplifying the production of child sexual abuse material, prompting a sharp rise in reports to watchdogs and law‑enforcement agencies. Reuters found actionable AI‑generated CSAM reports more than doubled over two years, while the Internet Watch Foundation logged 8,029 such images and videos in 2025 alone. The National Center for Missing & Exploited Children received 1.5 million AI‑linked reports that year, up from 67,000 the previous year. A Minnesota case involving a school employee illustrated how everyday photos can be weaponized, leaving investigators to untangle whether a child in an image is real, altered or entirely fabricated. Read more