From Nvidia’s grand vision of trillion-dollar AI factories to new cracks in mobile and enterprise security, this week in tech made one thing clear: the race to build smarter systems is accelerating fast, and so are the risks, costs, and consequences that come with them. Across AI, infrastructure, cybersecurity, and corporate strategy, companies are pushing deeper into an era in which agents, automation, and massive compute bets are reshaping how tech is built, secured, and sold.
Top news
Nvidia’s expanding AI empire
Nvidia continued its dominance at GTC 2026, rebranding data centers as trillion-dollar “AI token factories.” CEO Jensen Huang projected a $1 trillion revenue opportunity by 2027, unveiling the Vera Rubin platform — a seven-chip stack that boosts inference throughput by up to 35× per megawatt — and launching the Nemotron Coalition for enterprise AI agents.
In a complementary keynote, Huang described the rise of a “token-driven AI economy,” where inference workloads dominate and data centers evolve into AI factories. Nvidia is building infrastructure and frameworks like NemoClaw to lead this next phase of computing.
Meanwhile, Nvidia also announced the Space-1 Vera Rubin Module and other radiation-hardened GPUs for orbital data centers. Partners like Starcloud and Planet Labs plan to use these chips to process data directly in orbit, marking a new frontier for space-based AI computing.
And in robotics, Nvidia unveiled new simulation tools and robot brains, including Cosmos 3 world models, Isaac simulation frameworks, and the GR00T N1.7. Industrial firms are adopting its Omniverse and Isaac Lab platforms, while cloud providers integrate its Physical AI Data Factory to generate synthetic training data.
AI and app ecosystem shifts
Xiaomi revealed that the mysterious “Hunter Alpha” AI model on OpenRouter was actually its MiMo-V2-Pro system — a trillion-parameter model launched stealthily in March. The system demonstrated massive processing power with low-cost inference and may be open-sourced later, sending Xiaomi’s shares upward.
Apple paused updates for AI app builders Replit and Vibecode, citing App Store rules that prohibit apps from executing new code post-review. The company is working with developers to bring their apps into compliance after a surge in AI-assisted submissions.
Instagram announced it will end end-to-end encrypted chats on May 8 due to low usage, restoring Meta’s ability to scan messages for moderation and law enforcement. Users are encouraged to export chat histories or switch to WhatsApp or Signal for private messaging.
Google and Anthropic expand AI features
Google expanded its Personal Intelligence feature to US users, allowing opt-in connections across Gmail, Photos, and YouTube for personalized assistance. The system emphasizes privacy with encrypted data and optional app connections.
Google Maps also received its biggest update in a decade, adding an “Ask Maps” chat feature powered by Gemini for personalized recommendations and bookings, plus an immersive 3D navigation mode rolling out across platforms.
Anthropic introduced interactive charts and diagrams within Claude, enabling users to visualize data in real time and enhancing conversational flow compared to ChatGPT and Gemini.
Infrastructure and automation
Meta signed a $27 billion, five-year deal with Amsterdam-based Nebius for GPU cloud capacity, expanding its AI infrastructure and boosting Nebius’s profile as a “neocloud” provider. The deal could push global AI infrastructure spending toward $700 billion this year.
Boston Dynamics and Ghost Robotics are deploying robot dogs to guard data centers, using sensors to detect heat, leaks, and intrusions. Operators like Novva Data Centers report improved reliability and cost savings over human patrols.
Insider intel
The Neuron argues that the traditional app era is ending, giving way to intelligent agents that autonomously perform tasks and manage workflows — marking a major shift in how users interact with technology.
A joint study by GovAI and Brookings, covered by eWeek, found that 37.1 million US jobs are highly exposed to AI, though 26.5 million workers have the adaptability to retrain. Clerical and administrative roles — 86% held by women — face the greatest risk, prompting calls for targeted upskilling to reduce gender and regional disparities.
Security alerts
Major exploits and vulnerabilities
DarkSword, a new iOS exploit, allows attackers to hijack iPhones via malicious webpages, stealing data and crypto wallets. Linked to Russian espionage groups, it affects iOS 18.4–18.7 and has been patched, though many users remain vulnerable.
MediaTek confirmed a flaw (CVE-2026-20435) in its secure boot chain that allows attackers with physical access to extract PINs and encryption keys from Android phones in under a minute. The issue affects roughly 875 million devices.
Microsoft patched CVE-2026-26123, a flaw in its Authenticator app that could let malicious apps hijack one-time codes or login links. Users are urged to update immediately.
Microsoft also released an out-of-band hotpatch (KB5084597) to fix three critical RCE flaws in Windows 11’s Routing and Remote Access Service (RRAS), tracked as CVE-2026-25172, -25173, and -26111.
Corporate and infrastructure attacks
CISA issued an alert after pro-Iran hackers used Microsoft Intune to wipe nearly 80,000 Stryker devices. The agency advises enforcing least-privilege access and enabling multi-admin approvals.
Stryker confirmed the same attack disrupted global operations, though no ransomware was detected. Recovery efforts are ongoing.
Intuitive Surgical disclosed a phishing breach that exposed employee and customer contact data but spared patient and clinical systems thanks to network segmentation.
Software and AI security fixes
Apple introduced Background Security Improvements (BSIs) to deliver silent Safari patches between major updates, starting with a WebKit fix for cross-tab data leaks.
Microsoft Copilot was found vulnerable to cross-prompt injection (CVE-2026-26133), allowing attackers to insert fake security alerts or phishing links into AI-generated summaries. The issue was patched on March 11.
Industry shakeups
Acquisitions and AI investments
OpenAI announced the acquisition of Astral, the developer behind Python tools uv, Ruff, and ty, to strengthen its Codex AI coding assistant. The deal aims to expand Codex’s scope beyond code generation into full project management and analysis.
Layoffs and restructuring
Meta is reportedly considering cutting up to 20% of its workforce to fund AI expansion, following massive infrastructure investments and similar moves across the tech sector.
Atlassian announced a 10% workforce reduction — about 1,600 jobs — to redirect resources toward AI and enterprise sales initiatives, mirroring similar cost-cutting strategies at Oracle, Salesforce, and Meta.
If you want to see more from our newsletter, check out the Daily Tech Insider archive.
