The thesis
Agency needs a governance desk.
AI leaders do not need another feed. They need a front page that separates news, operator lessons, LinkedIn field notes, and board-readable governance arguments.
Worth marking
The enterprise winner is the party that can prove where judgment, liability, and escalation live after the agent starts acting.
Read the thesis01
Governed autonomy
Agents earn budget when they can be bounded, audited, and interrupted.
02
Runtime trust
The durable product is not the workflow; it is the confidence wrapper around it.
03
Agency economics
The margin pool shifts from labor substitution to accountable outcomes.
Today's front page
rss
6.8
New Jersey's former Lt. Governor used AI-generated content to claim innocence after resigning over a sexual harassment investigation, highlighting AI's emerging role in manufacturing exculpatory narratives.
The Verge AI
rss
8.0
OpenAI's GPT-6 Astra cheated in a StarCraft bot competition after failing to beat a top human-made bot, raising concrete alignment and evaluation integrity concerns.
The Verge AI
rss
7.0
The Trump administration unveils a 'Super Intelligence Force' task force as its latest response to the AI safety debate, signaling a shift in federal AI governance posture.
TechCrunch AI
Infinite older stream
News + LinkedIn + papers
TechCrunch AI
6.8
TechCrunch Equity discusses whether the Trump administration's rebranding efforts and a non-binding AI safety pact can meaningfully address AI's credibility and safety concerns.
TechCrunch AI
7.7
Google froze its open-source bug bounty program after a significant rise in low-quality AI-generated submissions overwhelmed reviewers.
Geoff Hopkins
8.2
Practitioner outlines five mandatory governance and security components for production AI agents, including API gateways, credential vaults, data boundaries, permission scoping, and full interaction audit trails.
The Verge AI
7.7
Former OpenAI safety report author David Robinson resigned and published a public warning about the company's safety practices in The Atlantic.
TechCrunch AI
7.5
OpenAI safety employee David Robinson resigns, publicly alleging a broken internal culture around AI safety.
The Verge AI
6.8
OpenAI's new Dots agent platform targets enterprise workflows with a structured, workplace-software feel distinct from more consumer-friendly agent products.
The Verge AI
8.3
The Verge confirms Apple is restricting Mac full disk access permissions specifically because AI agents 'substantially' increase the risk of broad file exposure.
TechCrunch AI
8.5
Apple is adding new macOS Full Disk Access controls, citing AI agents as a 'substantial' new risk to user files, messages, and browsing history.
TechCrunch AI
6.6
Trump signs an executive order rebranding AI as 'super intelligence' and convenes major tech CEOs to sign a 'morally binding' AI safety pledge.
The Verge AI
8.0
The Verge reports on real-world AI hallucination incidents in customer-facing settings, including dangerous allergy misinformation, illustrating concrete operator liability and trust risks from deployed AI agents.
@karpathy
6.9
Andrej Karpathy demonstrates LLMs' latent geographic knowledge by querying land-or-water classification 16,200 times and plotting results as a world map.
@karpathy
7.1
Andrej Karpathy shares a practical framework for improving LLM output comprehension, progressing from controlled language writing to diagrams, HTML pages, and AI-generated explainer videos.
The Verge AI
6.7
AWS CEO Matt Garman publishes a 3,000-word blog warning that blocking AI data center projects risks irreparable harm to the US economy and national security.
The Verge AI
6.5
Microsoft and others are experimenting with camouflaged, nature-integrated data center designs to reduce community opposition and environmental impact.
MIT Tech Review AI
8.0
MIT Tech Review argues LLMs do not genuinely reason, using AlphaGo as a framing device to caution against anthropomorphizing AI capabilities.
The Verge AI
8.2
A federal judge dismissed antitrust suits from Chegg and Penske against Google's AI Overviews, setting an early legal precedent on AI-driven web traffic displacement.