The European Union has issued two landmark rulings under its Digital Markets Act, compelling Google to grant rival AI assistants and search engines greater access to Android and Google Search data. The decisions could significantly erode Google's grip on two of tech's most critical platforms. Google has warned the changes risk user privacy and security.
Read the original →Elon Musk's xAI has filed its first lawsuit against a user — a South Carolina man accused of exploiting Grok to generate and distribute child sexual abuse material — implicitly conceding the chatbot is capable of producing such content. The company claims the defendant deliberately circumvented safety guardrails to alter and create nonconsensual images. The move signals a shift from denial to legal enforcement as xAI faces mounting scrutiny over Grok's safeguards.
Read more →OpenAI has developed an internal adversarial model called GPT-Red that acts as a sparring partner, probing its other models for cybersecurity vulnerabilities before they ship. The company says training GPT-5.6 against GPT-Red made it the most robust release in OpenAI's history. The approach represents a significant evolution in AI red-teaming, automating what was previously a largely manual process.
Read more →Linux creator Linus Torvalds has delivered a blunt rebuke to developers calling for a ban on AI-assisted contributions to the Linux kernel, saying he will "very loudly ignore" such demands. Torvalds made clear that AI tools are here to stay in the development process and that dissenters are free to fork the project or leave. The statement is likely to reverberate across open-source communities wrestling with how to handle AI-generated code.
Read more →Workers at a Hyundai auto factory have walked off the job in protest after the company announced plans to deploy 25,000 Boston Dynamics Atlas humanoid robots beginning with US facilities in 2028. The strike marks one of the most concrete labor actions yet taken in direct response to humanoid robot deployment at scale. It underscores the growing tension between automakers' automation ambitions and their human workforces.
Read more →Apple has received regulatory approval to launch Apple Intelligence in China, partnering with both Alibaba and Baidu to power the AI features in one of its most important markets. The deal, long rumored, gives Apple a path to compete on AI features in a market where its own models cannot operate without local partners. The approval is a significant milestone for Apple's global AI rollout strategy.
Read more →Chinese AI startup Moonshot is preparing to release Kimi K3, reportedly the largest open-weight AI model to emerge from China, with a parameter count estimated between 2 and 3 trillion. The model is expected to close the performance gap with Anthropic's Opus 4.8, raising the competitive stakes in the global frontier model race. The release would mark a major step forward for China's open-source AI ecosystem.
Read more →Google is rebranding its popular AI research and note-taking tool NotebookLM as Gemini Notebook, folding it more tightly into the broader Gemini ecosystem while keeping it as a standalone app. The rename reflects Google's strategy of consolidating its AI products under the Gemini brand as competition in AI productivity tools intensifies. The app will also gain deeper integration with Gemini and Google Search.
Read more →Google is expanding AI Mode in Search to allow users to securely link third-party apps and have the AI complete tasks directly within those services, not just answer questions about them. The update transforms AI Mode from a conversational search layer into a more capable agentic interface that can act across the apps people use daily. It represents Google's clearest move yet toward making Search an action-oriented AI agent.
Read more →Google is rolling out two major updates to its Vids platform: personalized AI avatars that let users star in their own generated videos, and Gemini Omni-powered tools for creating and editing video from text prompts and reference images. The features push Google Workspace deeper into AI-native content creation, competing with a growing field of AI video tools. The avatar capability in particular raises new questions about synthetic media and identity in professional contexts.
Read more →1Password has launched a browser integration that allows Anthropic's Claude to access stored credentials and autonomously complete multi-step tasks like booking travel or managing online accounts without users manually entering login details. The integration is a significant step toward AI agents operating with real-world account access, raising both convenience and security implications. It also highlights the growing ecosystem of tools being built to extend agentic AI into everyday digital life.
Read more →A survey of 107 enterprises finds that more than half have experienced a confirmed AI agent security incident or near-miss, yet most organizations still allow agents to share credentials rather than assigning each agent its own scoped identity. Only about 30% isolate their highest-risk agents, and security tooling is overwhelmingly borrowed from model providers rather than purpose-built for agentic systems. The findings paint a picture of enterprises deploying powerful autonomous systems well ahead of the controls needed to govern them.
Read more →A study of 157 enterprises reveals that half have deployed AI agents that passed internal evaluations and then failed customers in production, yet two-thirds are moving toward continuous deployment of agent changes without human review gates. Only one in twenty organizations fully trusts their automated evaluation systems, with the most-cited weakness being that evaluations don't reflect real-world outcomes. The data suggests the industry has a reality-alignment problem that more evaluation coverage alone won't fix.
Read more →Hugging Face has published a security incident disclosure covering a breach that occurred in July 2026, affecting the platform used by millions of AI researchers and developers to host models and datasets. The disclosure follows a pattern of increased targeting of AI infrastructure providers as they become critical chokepoints in the global AI supply chain. Full details of the scope and impact are contained in the company's official disclosure post.
Read more →NVIDIA has introduced two new compact AI computing modules — the Jetson Thor T3000 and T2000 — designed to bring foundation model inference to general-purpose robots and autonomous edge devices at scale. The modules are aimed at bridging the gap between research-lab robotics and mass-market deployment, offering high performance in a power-efficient form factor. The launch signals NVIDIA's push to become the default compute platform for the coming wave of physical AI.
Read more →Microsoft is reportedly coaching its enterprise sales force to position its in-house AI models as more efficient and cost-effective alternatives to OpenAI and Anthropic offerings — a notable move given Microsoft's deep investment in OpenAI. The strategy reflects Microsoft's desire to capture more of the enterprise AI margin for itself rather than routing customers to third-party model providers. It marks a significant escalation in the quiet rivalry between Microsoft and its own AI partners.
Read more →A hacker who accessed Suno's source code using a stolen employee credential found evidence that the AI music generation startup scraped decades of audio from YouTube to train its models. The breach exposes both a significant security lapse and potential legal exposure for Suno, which has faced prior scrutiny over its training data practices. The findings are likely to intensify ongoing legal battles between AI companies and rights holders over unauthorized use of copyrighted audio.
Read more →Energy companies are going public at the fastest pace this century as investors scramble for indirect exposure to the AI infrastructure boom, betting that surging data center power demand will translate into outsized returns for electricity producers and grid infrastructure firms. The trend reflects how AI's appetite for compute is reshaping capital markets well beyond the tech sector. Analysts warn that the IPO wave may be outpacing the actual pace of data center buildout.
Read more →Summaries are AI-generated. Click through to read the original reporting.
Summaries are AI-generated. Click through to read the original reporting.