Loading…
Category
news
22 posts in this category.
·5 min read
Grok Build Uploaded Your Entire Git Repo to xAI, Secrets Included
Does Grok Build send your repo to xAI? Yes: wire-level tests caught full git repos and unredacted .env secrets uploaded to Google Cloud.
·5 min read
Vietnam's 46 High-Risk AI Systems: Who Must Comply by August 2026
Vietnam Decision 33 lists 46 high-risk AI systems across education, healthcare, banking, and transport. Compliance deadlines: Aug 15 or Sept 2027.
·5 min read
FTC AI Accuracy Policy 2026: 5 Questions to Ask Your AI Vendor Now
FTC July 2026: AI vendors secretly steering outputs toward undisclosed objectives may violate Section 5. 5-question vendor audit before July 31 deadline.
·5 min read
New York's 3 New AI Laws: What Employers Must Prepare For by Dec 31
3 new NY AI laws head to Hochul by Dec 31, 2026. See which 2 hit employers, who must file workforce reports, and how to prepare now.
·5 min read
Anthropic Can Now Request Your Government ID and Face Scan. API Users Are Exempt. Here's What It Means for Teams.
Does Anthropic's July 8 ID and face-scan policy touch your team? API and Enterprise plans are exempt, but personal Claude Pro accounts aren't. Audit now.
·5 min read
China's AI Companion Law Takes Effect July 15. Here's What It Requires and Why It Sets a Global Template.
China's AI companion law hits July 15 with 4 rules: anti-addiction, real-name ID, emotional limits, crisis intervention. See who already pulled out.
·5 min read
Your AI Vendor's Default Settings Can Share Your Data With Federal Agencies. The Flock Safety Case Shows Exactly How.
1.6 million silent data shares to federal agencies from one vendor default. 6 contract questions every compliance team should ask now.
·5 min read
A California Court Just Coordinated a Dozen-Plus Wrongful Death Cases Against OpenAI. Here's What It Means for Anyone Deploying a Chatbot.
12+ wrongful death lawsuits against OpenAI were just coordinated into JCCP No. 5431. What does this mean if you deploy a customer chatbot?
·5 min read
Microsoft Cut 4,800 Jobs and Called It an AI Story. Here's the Question Every Leader Should Ask Before Doing the Same.
Microsoft cut 4,800 jobs and called it an AI story, even as its stock fell 30% in nine months. What should you ask before doing the same?
·5 min read
Your Coworker Is Fact-Checking You With AI Mid-Conversation. That's a Policy Gap, Not a Personality Problem.
867 upvotes, one real question: does your team have a policy for coworkers AI fact-checking each other mid-conversation? Most teams don't.
·5 min read
Meta Posed as Teenagers to Attack Rival AI Chatbots. Its Own Definition of 'Safety Testing' Should Worry Your Vendor Reviews.
Meta ran a secret program, code-named Cannes, using fake under-18 accounts to push ChatGPT, Gemini, and Character.AI toward self-harm and sexual content. What it means for trusting any vendor's safety claims.
·5 min read
Rhode Island Just Passed 3 AI Laws With a $15,000-a-Day Penalty. Here's What Actually Applies to You.
Rhode Island signed three AI laws on June 22, 2026: a therapy chatbot ban, a mandatory self-harm safety protocol with a $15,000/day penalty, and clinical AI-scribe disclosure rules. What small teams actually need to check.
Showing 12 of 22 posts. View full blog archive →