The afternoon briefing.
The Trump administration eyes a stake in OpenAI, Meta seeks billions for AI, and Google commits to massive compute with SpaceX. Plus, Apple's WWDC plans and critical cybersecurity alerts.
AI Investment & Policy. The Trump administration is reportedly discussing a potential government stake in OpenAI, a move that could significantly reshape the AI startup landscape and its relationship with federal power. This comes as Meta also explores raising tens of billions of dollars through a stock offering to fuel its own aggressive AI initiatives. These developments underscore the immense capital flowing into AI and the growing government interest in its strategic control. Meanwhile, Florida has filed a lawsuit against OpenAI and CEO Sam Altman, treating ChatGPT as a defective product.
AI Infrastructure & Optimization. Google has committed to a massive deal, agreeing to pay SpaceX $920 million per month for compute resources, highlighting the escalating demand for advanced infrastructure to power AI and other high-performance tasks. This substantial investment reflects the ongoing race to secure computational power. Concurrently, the industry is seeing a shift towards "model routing," a strategy to optimize AI spending by matching tasks to the most appropriate model rather than defaulting to the most powerful, which could impact revenue for large model providers like OpenAI and Anthropic.
On-Device AI & Ecosystems. As Apple's WWDC approaches, expectations are high for significant AI announcements, with iOS 27 rumored to bring new designs to iPhone apps and Touch ID potentially returning for the iPhone Ultra. This focus on integrating AI deeply into consumer devices is echoed by Google's release of Gemma 4 QAT models, optimized for mobile and laptop efficiency. These moves signal a broader industry push towards more powerful and efficient on-device AI capabilities, shaping the future of personal computing.
Cybersecurity & Data Breaches. The cybersecurity landscape remains active, with a particularly alarming warning from Google and the FBI about a ransomware group that sends fake IT workers to physically infiltrate victim organizations. This sophisticated social engineering tactic highlights evolving threats beyond traditional digital attacks. Separately, a significant data breach exposed 2.6 million DentaQuest accounts, underscoring the persistent vulnerability of personal health information. Additionally, a Cisco SD-WAN zero-day vulnerability is currently under active attack with no patch available.
Trump administration, OpenAI discussing possible government stake in the AI startup
OpenAI CEO Sam Altman reportedly discussed the idea of a government stake in the AI startup with the Trump administration in 2025. This potential move could give the US government a direct interest in a leading AI developer.
Meta's stock sinks on report company could raise tens of billions of dollars to fund AI push
Meta's stock declined after a report suggested the company might raise tens of billions of dollars through a stock offering to fund its ambitious AI development efforts. This potential capital raise indicates Meta's aggressive push into the AI sector.
Google will pay SpaceX $920M per month for compute
Google announced a deal to pay SpaceX $920 million per month for compute services, just ahead of SpaceX's anticipated IPO. This massive agreement highlights the growing demand for high-performance computing infrastructure.
Florida's lawsuit against OpenAI and CEO Altman treats ChatGPT as a defective product and public nuisance
Florida has filed a lawsuit against OpenAI and CEO Sam Altman, alleging that ChatGPT is a defective product and a public nuisance due to risks to minors and inadequate safety measures. The 83-page complaint seeks billions in penalties and could set a precedent for the chatbot industry.
Google and FBI warn of ransomware group that sends fake IT workers to hack victims in person
Google and the FBI have issued a warning about a ransomware group, Silent Ransom Group, that deploys fake IT workers to physically infiltrate law firms and steal data using USB drives or remote access tools. This new tactic represents a significant escalation in cyberattack methods.
Researchers used AI and known genetic codes to develop a ‘fundamentally new’ vaccine
Researchers have utilized AI and known genetic codes to develop a "fundamentally new" vaccine, marking a significant shift in pandemic preparedness. This innovative method aims to accelerate the creation of vaccines targeting entire families of viruses.
Model routing is a fix for AI overspending. That's a problem for OpenAI and Anthropic
Companies are increasingly adopting "model routing" to optimize AI spending by matching specific tasks to the most suitable AI model, rather than relying solely on the most powerful ones. This practice poses a challenge for companies like OpenAI and Anthropic, which primarily offer high-end models.
New iOS 27 designs reportedly coming to these iPhone apps
Rumors suggest that iOS 27, set to be unveiled in days, will introduce new design changes to various iPhone apps. These updates are expected to impact both major and minor applications within the Apple ecosystem.
Apple's WWDC: Tim Cook's AI legacy at stake in his final developer conference as CEO
Apple's upcoming WWDC is seen as a pivotal moment for CEO Tim Cook's AI legacy, with the future of Siri as an agentic platform and the company's stock valuation at stake. The conference is expected to reveal Apple's comprehensive strategy for integrating AI.
Yet another Cisco SD-WAN 0-day under attack, and no patch in sight
A new zero-day vulnerability in Cisco SD-WAN is currently under active attack, with no patch yet available for system administrators. This critical security flaw requires immediate attention from affected organizations.
2.6 million DentaQuest accounts exposed by data breach
DentaQuest confirmed a data breach that exposed 2.6 million accounts, with the ShinyHunters group claiming to have stolen 234GB of data after negotiations reportedly broke down. The incident highlights ongoing risks to sensitive personal information in healthcare.
Gemma 4 QAT models: Optimizing compression for mobile and laptop efficiency
Google has released Gemma 4 QAT models, specifically optimized for compression to enhance efficiency on mobile and laptop devices. This development aims to bring more powerful AI capabilities to edge devices.
Supabase doubles valuation to $10B in 8 months
Supabase, an open-source project turned company, has doubled its valuation to $10 billion in just eight months, benefiting significantly from AI tools like Claude and Codex. This rapid growth demonstrates the potential for open-source platforms leveraging AI.
"We pissed off a lot of people": Giant data center plan cut 50% amid protests
A developer's plan for a giant data center was cut by 50% following significant public protests, with the developer stating they felt "beaten up" and had "no choice" but to reduce the scale. This incident highlights community resistance to large-scale tech infrastructure projects.