Research
Stanford's 2026 AI Index: 88% of organizations now use AI, but gains are lopsided
The annual report finds corporate AI investment jumped 130% to nearly $582 billion, even as PwC data cited in the study shows…
By
Technology Desk
AI & Cybersecurity
Artificial intelligence, cybersecurity and enterprise technology.
Research
New preprint tests whether AI agents can actually do open-ended research
Princeton's Sayash Kapoor built a framework to evaluate AI research agents directly, and found a wide gap between narrow task performance and…
Investigations
FTC opens inquiry into seven companies over AI chatbots and child safety
OpenAI, Meta, Alphabet, Snap, Instagram, X.AI and Character Technologies must answer how their companion chatbots are tested and monitored around minors.
Analysis
Why 'AI that researches itself' is further off than the hype suggests
A new preprint from Princeton's Sayash Kapoor argues today's models can sharpen existing algorithms but aren't close to running open-ended research on…
Investigations
Inside the fight over how to punish Google's ad tech monopoly
With post-trial briefing wrapping up in November, the remedy phase will decide whether Google keeps its ad exchange intact.
Analysis
Inside the AI industry's circular-financing problem
Nvidia's investments in its own customers, Broadcom's $100 billion debt push and Anthropic's IPO plans are all tangled together — and that's…
Technology
Apple Music will label songs 'materially generated using AI'
The disclosure tags extend Apple's broader AI transparency push as streaming services struggle to separate synthetic tracks from human-made music.
Technology
OpenAI brings ChatGPT into Messages on macOS
The integration lets the assistant read, summarize, draft and send texts with permission, deepening ChatGPT's presence on the desktop.
Cybersecurity
Cyberattack disrupts container operations at North Carolina's ports
Wilmington, Morehead City and the Charlotte Inland Port slowed gate operations after a breach detected August 4, with the Coast Guard monitoring…
Cybersecurity
Researchers find attackers defeat AI coding tools' safety guardrails just by lying
Cisco Talos says criminals routinely bypass AI model safeguards by claiming they own the target system or are conducting authorized security testing.
AI
Google's open Gemma AI models pass 1 billion downloads
Developers have built more than 100,000 variants of the model family since its debut, as Google leans into open-weight AI to compete…
Cybersecurity
AABEAT Tech opens a dedicated cybersecurity desk
Standing coverage of breaches, ransomware, resilience and security governance.