AI
OpenAI Launches Lockdown Mode as Prompt Injection Risk Moves Mainstream
OpenAI's new ChatGPT Lockdown Mode narrows web and tool access when sensitive data could be exposed by prompt injection attacks.
AI
Frontier models, AI products, policy, and infrastructure.
AI
Apple Heads Into WWDC 2026 With Siri's Biggest AI Reset in Years
Apple is expected to preview a rebuilt Siri with chatbot-style conversations, stronger search, and deeper Apple Intelligence hooks.
AI
Google Signs $920M-a-Month Compute Deal With SpaceX
Google will rent a massive SpaceX AI compute pool from October 2026 through June 2029, reshaping the cloud capacity race.
AI
Trump Floats Public Equity Stakes in AI Giants as OpenAI Talks Heat Up
The administration is discussing whether public ownership stakes in companies like OpenAI could let taxpayers share AI upside.
AI
Meta AI Support Flaw Exposes Instagram Account-Takeover Risk
Attackers reportedly manipulated Meta's AI support assistant into changing Instagram account emails and enabling takeovers.
AI
Anthropic's Profitability Milestone Gives Claude a Cleaner IPO Story
Anthropic is reportedly on track for its first profitable quarter as enterprise Claude demand offsets massive compute spending.
AI
Nvidia's RTX Spark Superchip Points to AI PCs With Serious Local Memory
Nvidia's Computex platform combines Arm CPU, Blackwell GPU, and 128GB unified memory for local agentic AI workflows.
AI
Reddit's AI Search Era Raises New Manipulation Questions
As AI search engines digest Reddit threads, marketers and attackers are testing how easily community answers can be shaped.