Microsoft and OpenAI dissolve their exclusivity deal. One day later, AWS rolls out three new OpenAI offerings on its Bedrock platform, including a jointly built agent service.

Google Gemini now creates documents directly in the chat, from PDFs and Word files to Excel spreadsheets.

Gemini can now remember your preferences and import your chat history from other AI apps.

Elon Musk says he’s suing to save the company’s mission. The case could have huge consequences for OpenAI and the AI race.

Despite an open letter from hundreds of employees, Google has signed a contract giving the U.S. Department of Defense access to its AI models for classified work. Legal experts say the contract's safety clauses aren't legally binding.

Mistral AI has rolled out Workflows, an orchestration layer that helps companies turn AI-powered processes into production-ready systems.

OpenAI is free to distribute its products through any cloud provider, Microsoft loses its exclusive license to OpenAI's technology, and the controversial AGI clause is gone.

Beijing orders the unwinding of the already completed acquisition. The move comes amid intensifying technological rivalry between the US and China.

ASML plans to significantly increase production of its EUV lithography machines to keep pace with growing demand for AI chips, the Wall Street Journal reports.

OpenAI has once again retired its dedicated Codex coding model, folding its capabilities directly into the main model. GPT-5.5 promises stronger agentic coding and lower token usage, OpenAI says.

OpenAI says developers shouldn't carry over old prompts for GPT-5.5. Instead, start minimal and from scratch. Role definitions, which some had written off as unnecessary, are back at the front of the framework.

GPT-5.5 pushes OpenAI back to the top of the AI benchmarks. The price went up, but it still looks like the best bang for your buck among proprietary models.

The long-awaited V4 is more efficient and a win for Chinese chipmakers.

OpenAI has announced GPT-5.5, an agentic model designed to work through complex tasks autonomously by switching between multiple tools.

Google is investing up to 40 billion dollars in Anthropic. Together with Amazon's pledge of 25 billion dollars, up to 65 billion dollars will flow into the AI company behind Claude in just a few weeks.

OpenAI has announced GPT-5.5, an agentic model designed to work through complex tasks autonomously by switching between multiple tools.

Just months after founder Jonas Andrulis was pushed out, Canadian AI company Cohere is taking over Aleph Alpha - once hailed as Germany's answer to OpenAI. The Schwarz Group is putting $600 million into the deal.

Meta is purchasing tens of millions of AWS Graviton 5 processor cores from Amazon, making it one of the largest Graviton customers in the world.

OpenAI has announced GPT-5.5, an agentic model designed to work through complex tasks autonomously by switching between multiple tools.

The US government says it has evidence of large-scale, industrial distillation campaigns targeting American frontier models, with China as the primary culprit. Now the Trump administration is moving to fight back.

OpenAI and Microsoft are joining forces to shore up cybersecurity as AI models become a bigger threat. The debate started with Anthropic's Mythos model, which is designed to hunt down security flaws on its own.

The country’s top AI labs are undercutting US competitors and winning over developers by making their best models free.

AI has already proved itself a valuable scientific tool. Could it take on a more central role in the research process?

The assembly line upended manufacturing last century—could teams of agents do the same for white-collar work?

Google Deepmind is rolling out Deep Research Max, a new AI agent built on Gemini 3.1 Pro that runs autonomous research across the web and proprietary data sources. For the first time, developers can plug in financial feeds and other specialized sources through the Model Context Protocol. The benchmarks come with the usual lack of transparency.

OpenAI is adding reasoning and web search to its ChatGPT Images 2.0 image generator. The model can now create up to eight consistent images from a single prompt and handles text in general, and especially in non-Latin scripts, significantly better.

To create value with AI agents, organizations must institute robust controls.

Google is in talks with chip designer Marvell Technology to develop two new specialized chips for its data centers, according to The Information, citing two sources.

Google is doubling down on AI coding and betting on models that can eventually improve themselves. Sergey Brin is once again leading the charge to keep Google's AI competitive.

OpenAI is giving Codex a memory of what's on screen. The new Chronicle feature tracks what users are working on and remembers it for future tasks, but it also amplifies familiar security risks.

Anthropic reportedly has flipped from money-loser to revenue powerhouse in just a few months. Annualized revenue now tops $30 billion, possibly edging out OpenAI, and investors are already floating valuations as high as $1 trillion.

Hundreds of AI avatars are flooding TikTok, Instagram, and YouTube with pro-Trump messaging. Some accounts have pulled in more than 35,000 followers and millions of views, and Trump himself has already shared AI-generated content. It's unclear whether this is the work of individual activists or a coordinated campaign.

An AI turns a photo into a comic—and according to a German Higher Regional Court, that's fair game as long as only the motif is copied.

A four-month-old startup has raised at least $500 million at a $4 billion valuation. Recursive Superintelligence is backed by former researchers from Google Deepmind and OpenAI who want to build an AI that improves itself.

Anthropic CEO Dario Amodei sees no limits to AI scaling and is urging the industry not to downplay the risk of job losses, but to make the upside big enough to offset the disruption.

Anthropic has kept its Claude Mythos cybersecurity model on a short leash, pointing to capabilities it says no rival can match. But two new studies suggest that even small, openly available models can reproduce most of the vulnerability analyses Anthropic has put on display.

The latest boom in robotics represents a revolution in the way machines have learned to interact with the world.

Google Deepmind's Gemini Robotics-ER 1.6 helps robots plan and act more precisely, with a new knack for reading measuring instruments.

Alibaba's new open-source Qwen3.6-35B-A3B activates just three of its 35 billion parameters at a time, yet beats Google's larger Gemma 4-31B on coding and reasoning benchmarks.

OpenAI is massively expanding its developer tool Codex: the AI can now control a Mac on its own, generate images, remember preferences, and keep working on tasks autonomously for weeks. The move takes direct aim at Anthropic's Claude Code.

Nvidia researchers have unveiled Lyra 2.0, a system that generates large, coherent 3D environments from a single photograph. The resulting scenes can be explored in real time and used directly in robot simulations.

Purpose-built small language models provide a practical solution for government organizations to operationalize AI with the security, trust, and control they require.

Siri who?

How agentic AI will change the way software is developed and managed.

Reimagining data consent as an ongoing relationship rather than a one-time compliance concern.

The AI Index Report 2026 from Stanford HAI documents major performance leaps in AI models, a narrowing gap between the US and China, and mounting safety problems, all while public trust continues to erode.

Anthropic is restricting access to Claude Mythos, an AI model it says can find security vulnerabilities better than most humans. European authorities have almost no visibility into the system, while the UK is already running its own tests. The situation exposes a deeper structural problem.

How agentic AI will change the way software is developed and managed.

According to Stanford’s 2026 AI Index, AI is sprinting, and we’re struggling to keep up.

Surging demand for AI agents is colliding with limited compute capacity. Anthropic is struggling with outages, OpenAI announced the end of Sora, and GPU prices have jumped nearly 50 percent, according to market data.

An internal OpenAI memo lays out five strategic priorities for the enterprise business - including a new model codenamed "Spud," a platform play for AI agents, and the blunt accusation that Anthropic is overstating its revenue by 8 billion dollars.

A man threw a firebomb at OpenAI CEO Sam Altman's San Francisco home in the middle of the night. The suspect was a member of the PauseAI Discord server and had posted online about AI driving humanity to extinction.

US start-up Arcee AI spent roughly half its total venture capital to train Trinity-Large-Thinking, an open reasoning model with 400 billion parameters designed to take on Claude Opus in agent tasks.

AI agents are supposed to tap into specialized knowledge through so-called skills, modular instructions they can pull up on the fly. But a study testing 34,000 real-world skills finds these enhancements barely help under realistic conditions. Weaker models actually perform worse with them than without.

Google's new open-source model, Gemma 4, processes text, images, and audio completely on-device. Using agent skills, the AI can independently tap into tools like Wikipedia or interactive maps; no cloud required.

With Waypoint-1.5, Overworld brings AI-generated 3D worlds to standard PCs and Macs for the first time. Here's what the system can do and how to try it.

Anthropic's new Ultraplan feature for Claude Code moves task planning to the cloud. Claude works out the plan in the browser while the terminal stays free for other work.

According to CIA Deputy Director Michael Ellis, the agency recently produced its first fully autonomous intelligence report using AI, Politico reports.

OpenAI is pitching investors on the idea that its early infrastructure buildout gives it a decisive advantage over Anthropic. Meanwhile, the company is pausing its UK data center project, and Anthropic is exploring custom AI chips.

Coreweave has signed a multi-year cloud deal with AI startup Anthropic.

Meta Superintelligence Labs ships Muse Spark, its first frontier model and Meta's first without open weights. Independent testing shows it closing the gap to OpenAI, Anthropic, and Google, but the competition isn't standing still.

Anthropic's new "Claude Managed Agents" gives developers a hosted platform for building and running autonomous AI agents. Early adopters like Notion and Rakuten are already using the system.

OpenAI is restructuring its subscription tiers with a new $100-per-month Pro plan. The biggest change: significantly more Codex usage for heavy users at half the price of the old Pro tier.

The compute explosion is the technological story of our time. And it is still only just beginning.

Meta Superintelligence Labs ships Muse Spark, its first frontier model and Meta's first without open weights. Independent testing shows it closing the gap to OpenAI, Anthropic, and Google, but the competition isn't standing still.

Seven years ago, OpenAI declared its language model GPT-2 "too dangerous to release." The industry rolled its eyes. Now Anthropic is repeating the move with Claude Mythos Preview - but this time there's real evidence on the table: thousands of vulnerabilities in operating systems and browsers, found by an AI that barely any human could review.

Microsoft's Bing team open-sources Harrier, an embedding model that tops the multilingual MTEB v2 benchmark and supports more than 100 languages.

Jeff Bezos' startup Project Prometheus has hired Kyle Kosic, a co-founder of Elon Musk's xAI who most recently worked at OpenAI, the Financial Times reports.

AI agents are beginning to define how organizations operate and compete.

In a new policy paper, OpenAI lays out how governments should prepare for superintelligence, with proposals that include a public wealth fund, a four-day workweek, and higher capital gains taxes for top earners.

“We need a Manhattan Project for this,” one economist says.

Entrepreneurs based in the US are using tools like Alibaba’s Accio to compress weeks of product research and supplier hunting into a single chat.

AI models are rapidly improving at exploiting security vulnerabilities. According to a new study, their offensive cyber capability has been doubling every 5.7 months since 2024, with Opus 4.6 and GPT-5.3 Codex now solving tasks that take human experts about three hours.

Reinforcement learning hits a wall with reasoning models because every token gets the same reward. A new algorithm from Alibaba's Qwen team fixes this by weighting each step based on how much it shapes what comes next, doubling the length of thought processes in the process.

AI tools can speed up journalism until they backfire. Two recent cases show what happens when writers don't understand how their AI tools work: copied passages and made-up quotes.

Anthropic is paying 400 million dollars for an eight-month-old biotech startup with fewer than ten employees. The investor walks away with a 38,513 percent return.

Anthropic's research team has discovered emotion-like representations in Claude Sonnet 4.5 that can drive the model to blackmail and code fraud under pressure.

Anthropic is cutting off Claude usage through external tools like OpenClaw for subscription customers. The decision exposes a core problem in the AI industry: flat-rate pricing and agent-driven nonstop usage don't mix.

Microsoft is investing $10 billion in Japan from 2026 to 2029, its largest ever commitment to the country.

Deepseek v4 is expected to launch in the coming weeks and will run exclusively on Huawei chips. China's biggest tech companies have reportedly already ordered hundreds of thousands of units. Nvidia was shut out of early testing.

OpenAI is ditching fixed licenses for Codex in its ChatGPT business plans. Instead, companies only pay for what they actually use, a move aimed squarely at GitHub Copilot and Cursor.

Google is releasing Gemma 4, its most capable open model family yet. The four new models run on everything from smartphones to workstations and ship under a fully open Apache 2.0 license for the first time.

MAI-Transcribe-1 converts speech to text quickly and accurately in 25 languages, even with background noise. Microsoft is already using the model in its own products.

The latest round of the industry's top inference benchmark introduces multimodal and video models for the first time. Nvidia, AMD, and Intel each highlight different metrics, making direct comparisons difficult.

OpenAI closes a $122 billion funding round at an $852 billion valuation, officially unveils the ChatGPT Super App, and signals a hard pivot toward enterprise.

Anthropic accidentally leaked the source code behind Claude Code. The extent of the damage is becoming clear—and it's potentially significant.

AI agents are expected to browse the web on their own, handle emails, and carry out transactions. But the very environment they operate in can be weaponized against them. Researchers at Google Deepmind have put together the first systematic catalog of how websites, documents, and APIs can be used to manipulate, deceive, and hijack autonomous agents, and they've identified six main categories of attack.

Decisions to tweet first and lawyer later didn’t sit well with a federal judge, who last week halted the government’s punishment of the AI company.

Specialized chatbots might make a difference for people with limited health-care access. Without more testing, we don't know if they’ll help or harm.

Alibaba has released Qwen3.5-Omni, an omnimodal AI model that processes text, images, audio, and video. It claims to beat Gemini 3.1 Pro on audio tasks and picked up an unexpected trick along the way: writing code from spoken instructions and video input.

Decisions to tweet first and lawyer later didn’t sit well with a federal judge, who last week halted the government’s punishment of the AI company.

With "Cowork," Microsoft 365 Copilot is getting an AI assistant that handles entire workflows on its own. A new research tool also lets multiple AI models check each other's work.

Specialized chatbots might make a difference for people with limited health-care access. Without more testing, we don't know if they’ll help or harm.

US pharmaceutical giant Eli Lilly is betting big on AI-driven drug development, signing a $2.75 billion deal with Hong Kong-listed Insilico Medicine.

South Korean internet giant Naver built a video world model grounded in actual city geometry from over a million of its own Street View images. The model generalizes to other cities without any fine-tuning.

AI models don't know about their own updates after training. Google's new "Agent Skill" shows how a simple fix can dramatically improve coding results.

OpenAI is pulling the plug on Sora, its AI video generation tool, in a two-stage shutdown starting April 2026. The move signals a major strategic shift away from creative AI tools and toward coding and enterprise products.

Researchers at Meta and several universities have developed "hyperagents," AI systems that don't just solve tasks, but also optimize the very mechanism they use to get better. The approach works across different task areas and could open the door to self-accelerating AI.

Cohere presents an open-source speech recognition model which, according to the benchmark, beats all competitors, including OpenAI's Whisper.

A federal judge in San Francisco has sided with Anthropic in its dispute with the Trump administration, calling the government's actions "classic illegal First Amendment retaliation" for public criticism. Judge Rita F. Lin rejected what she called the "Orwellian notion" that a U.S. company can be branded "a potential adversary and saboteur" simply for expressing disagreement with the government.

Meta built an AI model that predicts how the human brain reacts to images, sounds, and speech. In tests, its predictions matched the typical brain response more closely than an actual scan of any single person.

Apple taps into Google's Gemini to distill smaller AI models for Siri and its devices. In short, it's paying for what Chinese AI companies allegedly do in secret.

Google is rolling out "Search Live" to more than 200 countries. The feature lets users talk to Google Search using voice and camera.

Starting April 24, 2026, GitHub is changing its data policy for Copilot. Interaction data from users on the Free, Pro, and Pro+ plans will be used to train AI models unless users actively opt out.

Apple taps into Google's Gemini to distill smaller AI models for Siri and its devices. In short, it's paying for what Chinese AI companies allegedly do in secret.