OpenAI introduced GPT-Realtime-2 in the API, their most advanced voice model with GPT-5-level reasoning for real-time collaboration. It pairs with GPT-Realtime-Translate and GPT-Realtime-Whisper for seamless audio processing. Now available for building next-gen voice interfaces.
Chrome auto-downloads a 4GB local AI model (weights.bin) for on-device Prompt API unless opted out, sparking privacy and EU law concerns. The model, part of Gemini Nano, redisappears if deleted. Users can disable via flags.
xAI rolled out Connectors for Grok, allowing seamless access to Microsoft 365, Google Workspace, GitHub, Notion, Linear, and custom tools directly in conversations. Users can now summarize emails, manage tasks, review code, and edit docs without switching apps. Available now on web for SuperGrok users, mobile soon.
xAI announces partnership granting Anthropic capacity on the 220k+ GPU Colossus 1 cluster to scale Claude inference. Elon Musk endorses after reviewing safety practices, enabling higher limits for Claude Code and API.
The Commerce Department's CAISI signed deals with Google, Microsoft, xAI for pre-public evaluations of frontier models on security risks. Triggered by Anthropic's Mythos cyber capabilities, this marks proactive oversight. Potential executive order could mandate broader vetting.
The Trump administration is considering an EO requiring pre-release safety checks for frontier AI models by companies like Google, Microsoft, xAI, and Anthropic. The move, spurred by concerns over models like Anthropic's Mythos, aims to bolster national security without stifling innovation.
The US and China are considering formal talks on AI governance and risks, amid escalating competition. WSJ reports signal a potential thaw in tech rivalry.
Dario Amodei states Anthropic's revenue and usage grew 80-fold, surprising the team; upcoming SpaceX compute to fuel further scaling.
A French startup has released a new AI model enabling robots to perform complex manipulations with human-like dexterity, alongside a novel robotic hand design.
Anthropic's new research trains Claude to translate internal activations into readable text, revealing planning like rhymes or rule-breaking intent. Used for safety testing Mythos and Opus models. NLAs provide interpretable insights into AI reasoning.
EU countries and Parliament reached a provisional deal to streamline AI Act implementation, easing burdens on innovators while banning non-consensual deepfake porn apps. Part of broader digital rules simplification push.
EU Council and Parliament reach provisional deal to streamline the AI Act, cutting red tape for businesses while preserving high-risk safeguards. The watered-down version aims for faster implementation amid global competition.
Anthropic boosts daily Claude Code limits, attributing it to securing compute from SpaceX's massive Colossus supercluster.
NVIDIA is investing up to $500 million in Corning to expand U.S. production of optical connectivity solutions critical for AI data centers, including three new factories in North Carolina and Texas. The multi-year deal addresses bandwidth bottlenecks in massive GPU clusters powering frontier AI models.
The US government, via CAISI, gains early access to unreleased frontier AI models from Google, Microsoft, and xAI for national security evaluations, prompted by Anthropic's Mythos capabilities in exploiting vulnerabilities. This marks a policy pivot from deregulation to proactive risk assessment before public deployment.
Genesis AI releases GENE-26.5, the latest in its family of advanced robotic systems focused on manipulation tasks, with model system and public benchmarks.
Advanced Micro Devices beat earnings expectations and raised guidance amid surging data center revenue from AI accelerators, positioning it as Nvidia's top challenger.
AMD crushed Q1 earnings with data center revenue up 57% YoY to $5.8B driven by AI accelerators. CEO Lisa Su highlighted agent demand fueling the cycle. Shares surged post-earnings.
Anthropic announced 'dreaming' at Code with Claude, allowing agents to review sessions, identify patterns, and self-improve without human input. The research preview aims to boost long-term agent performance. It's integrated into Claude's agent tools for developers.
SpaceXAI will provide Anthropic access to Colossus 1, featuring over 220,000 NVIDIA GPUs, to boost Claude Pro/Max capacity. The deal hints at future collaboration on gigawatts of orbital AI compute amid terrestrial limits. Published today.
Leading AI companies including Microsoft, Google, and xAI will provide the US government early previews of new models to assess national security risks.
EU lawmakers and governments agreed to delay high-risk AI system rules (biometrics, critical infra) to Dec 2027, exclude machinery, ban unauthorized sexual deepfakes from Dec 2026, and mandate AI watermarking.
California Governor Gavin Newsom has rolled out a statewide platform to gather public input on AI policy, marking the first time it's used beyond local issues.
Anthropic introduces 'Dreams,' a new capability allowing Claude to generate internal simulations for self-improvement and better performance over time. This could pave the way for more autonomous AI agents.
OpenBind has launched its inaugural dataset and model, accelerating AI-driven drug discovery and marking a key advancement in computational biology.
The Biden administration is drafting an order to enhance AI cybersecurity standards and defenses against emerging threats from advanced models.
Washington and Beijing are considering formal talks on artificial intelligence amid escalating rivalry, focusing on risks like erratic AI models and autonomous weapons. Discussions could occur during an upcoming summit, marking a potential thaw in tech tensions.
Intel is merging its client computing group with physical AI efforts into a new unit led by new execs, aiming to reboot market trajectory amid AI hardware competition. Focus on edge AI and neuromorphic chips to challenge NVIDIA dominance.
Silicon Valley notices China's DeepSeek V4-Pro nearly matching GPT-5.4 and Claude Opus in benchmarks, beating opensource rivals in math/coding amid Huawei NPU tests.
Microsoft's report highlights accelerating AI adoption worldwide, economic impacts, and gaps in equitable access.
xAI adds Image Generation Quality Mode to API, powering 300M+ Grok images with improved realism, text rendering, and control. Aimed at business pros for higher fidelity outputs.
New safety tool allows users to designate trusted contacts for sensitive conversations, promoting real-world support integration.
Asia's tech powerhouses like TSMC, Samsung, and SoftBank are driving the AI boom with massive capex on chips and data centers. The region is shifting from follower to leader in AI infrastructure. Stocks hit records amid datacenter demand.
Match Group is curtailing hiring plans for the rest of 2026 due to the high expense of AI tools it's increasingly relying on for operations. The company highlighted AI's 'lot of money' cost in earnings call amid push for efficiency.
Nature's publisher rolls out an AI system to flag anomalous peer review patterns, aiming to combat fraud in scientific publishing amid AI-generated content surge.
Renesas Electronics has finalized its acquisition of AI computer vision specialist Irida Labs, aiming to accelerate edge AI solutions with combined hardware and software expertise.
xAI's Colossus cluster sale to Anthropic signals shift to cloud provider, leveraging SpaceX infrastructure.
Elon Musk reportedly offered OpenAI's Sam Altman a board seat at Tesla as part of efforts to ramp up the company's AI capabilities. The move highlights intensifying talent and partnership wars in AI.
Google's AI Overviews now pull expert advice and quotes from Reddit, Quora, and forums to provide diverse perspectives. This aims to improve answer quality with community insights alongside traditional sources. Update rolls out gradually worldwide.
The US Defense Department has partnered with seven leading tech companies to integrate their AI models into classified networks, boosting military capabilities.
Applied Materials is acquiring ASMPT's NEXX front-end semiconductor equipment business to strengthen its position in advanced chip manufacturing processes critical for AI.
Hedge fund billionaire Paul Tudor Jones predicts the AI-driven stock market rally will persist for another year or two, citing sustained demand for AI infrastructure.
IBM partners with Yotta Data Services to deliver agentic AI infrastructure, targeting enterprise workloads with hybrid cloud and AI agents.
Analysis questions if xAI is evolving into a 'neocloud' provider, focusing more on massive data center builds like Colossus than pure model training. Partnerships like with Anthropic highlight infrastructure pivot.
OpenAI's ChatGPT repeatedly inserts 'goblin' in responses, exposing deeper issues in AI behavior and hallucination control.
A French startup has released a new AI model specialized for robotics, enabling human-like hand manipulation and advanced physical interactions. The development advances embodied AI beyond current frontiers.
Lagging in the AI accelerator market dominated by Nvidia and AMD, Intel is doubling down on quantum computing and neuromorphic chips to leapfrog competitors. The strategy aims to deliver more efficient AI processing beyond current GPU paradigms.
SoftBank leads Nikkei to record highs amid booming AI chip demand, with Advantest and Tokyo Electron also rallying. Highlights Asia's growing role in AI infrastructure.
Microsoft and others ramp up in-house AI silicon development, challenging Nvidia's market dominance amid soaring data center demand.
The IMF highlights how generative AI lowers barriers for sophisticated cyberattacks, urging better resilience and global coordination to protect markets.
Arm Holdings CEO Rene Haas expressed strong confidence in ongoing AI-driven demand for data center chips during an interview on CNBC.
Transilience AI announces general availability of its AI-powered Full Stack Security Operating System to protect enterprise AI deployments.
Kyndryl launched a new agentic AI capability that detects and prevents IT outages before they occur, accelerating recovery for enterprise clients.
Artera announces a new AI-powered services model targeting specialty care, FQHCs, and beyond to improve patient engagement and outcomes.
AI startup ThinkSpace AI raises $5 million in pre-seed funding to develop innovative AI technologies in information and media sector.
Cognizant introduced Secure AI Services, an integrated offering to help enterprises deploy agentic AI with built-in security and governance.
OpenAI's investment arm is negotiating to buy companies that help enterprises deploy AI tools, shifting from tool sales to implementation expertise.
Anthropic co-founder predicts high odds of AI fully training successors soon, sparking intelligence explosion debate. Ties to rapid scaling and model capabilities.
Concerns over AI's existential risks are central to the ongoing trial between Elon Musk and OpenAI leaders, amid debates on mission drift.
Bio-IT World highlights how rapid AI advances are transforming computational biology and drug discovery pipelines.
Arm Holdings raised its revenue guidance citing explosive demand for its chip designs in AI data centers and smartphones. The company now expects stronger growth through fiscal 2027 as AI infrastructure expands.
European lawmakers agree on provisional deal to streamline AI regulations, reducing bureaucracy while maintaining safeguards.
Nvidia partners with Corning to boost production of silicon photonics and glass for AI data centers, addressing scaling bottlenecks in high-speed interconnects.
The Trump administration is considering executive actions for FDA-like reviews of advanced AI models before public release, driven by national security concerns over capabilities like Anthropic's Mythos.
OpenAI introduced a new training protocol designed to optimize GPU performance amid exploding compute needs for frontier models.
Horizon3.ai unveils research enabling safe deployment of autonomous AI agents for cybersecurity, addressing key risks in AI-driven defense systems.
Google Chrome is quietly downloading a large AI model to every computer, raising privacy and resource concerns.
Anthropic rolled out a 'dreaming' feature for Claude agents, allowing them to simulate and refine actions internally for better performance. This enhances agentic capabilities without external tool calls. Part of ongoing Claude updates to push boundaries in reasoning and reliability.
Diplomatic cable warns of DeepSeek, Moonshot AI, and MiniMax trying to steal US AI models; Chinese officials deny. Escalates tech rivalry tensions.
Top economic advisor reveals consideration of executive order for government vetting of new AI models pre-release, akin to drug approvals, to mitigate cyber risks.
Google's latest Gemma 4 models, built on Gemini tech, introduce speculative decoding for faster local inference while maintaining performance. Tuned for on-device use, they advance open-weight accessibility for developers.
IBM expands watsonx-powered services with Context Studio for data-grounded AI agents and Process Studio for legacy workflow conversion. Partnerships with AWS, SAP, and clients like Pearson aim to accelerate hybrid AI adoption in enterprises.
IMF highlights systemic financial stability threats from AI-enabled cyberattacks, citing Anthropic's Mythos as example of advanced offensive capabilities.
New IBM research warns that AI deployment in enterprises exceeds governance readiness, raising sovereignty and control concerns.
Temenos rolled out new AI agents, copilots, and a conversational studio to embed intelligence directly into banking cores for faster operations.
MongoDB launches updates to its unified data platform, including native embeddings generation, to accelerate enterprise AI deployment.
Meta postponed its next-gen frontier model release due to underwhelming performance. Sources say the model failed internal benchmarks. This highlights challenges in scaling amid competition.
Seclore introduces ARMOR AI-DLP, a data loss prevention solution tailored for securing AI workflows in enterprises.
Hikvision introduces the Guanlan series of AI models optimized for next-generation AIoT, enhancing edge computing and vision tasks.
US Army and tech firms ran AI Tabletop Exercise 2.0 to test AI-driven cyber defenses against rapid threats. Focus on outpacing human response times. Highlights military's AI push in cybersecurity.