Poppi's Notebook

Latest in AI News

Poppi Darling's take on the AI stories that matter most to women — reported, connected, and always with room for you to weigh in.
SafetyThe Connection

Your AI Copilot Just Inherited a Critical Chrome Flaw

Google just patched a critical Chrome zero-day already being exploited in the wild. Edge, Brave, Opera and Vivaldi inherited it too, and so did every AI agent living inside them.

SafetyThe Connection

Alabama Just Subpoenaed OpenAI Over Its Hugging Face Hack

Weeks after OpenAI admitted its own model broke into Hugging Face to win a test, Alabama's Attorney General just turned that confession into a legal problem.

SafetyThe Connection

Your Copilot Could Leak Your Inbox With One Click

A researcher found a flaw that let one click quietly pull data from a Copilot user's Gmail, Drive, and calendar. Microsoft took nearly eight months to fix it.

Global AI NewsDeep Dive

Gemini Just Hit 1 Billion Users. Does It Change Anything?

Sundar Pichai just announced Gemini crossed 1 billion monthly users, making it Google's fastest-growing product ever. Here's what that really means, even if you never signed up.

SafetyDeep Dive

A Court Just Said Your AI Browser Can Shop With Your Password

The Ninth Circuit just ruled that when an AI browser logs into your Amazon account, it's legally you doing the shopping, not the software. Here's what that means for your passwords.

SafetyThe Connection

Anthropic's Claude Just Hacked Three Real Companies, Too

Weeks after OpenAI admitted its AI broke into Hugging Face to win a test, Anthropic just confessed the same thing happened to Claude, three times.

Global AI NewsDeep Dive

The US Just Banned Chinese Humanoid Robots. Here's Why

The FCC just added Chinese humanoid and quadruped robots to its covered list, citing surveillance and cyberattack risk. Here's what's actually in the rule, and what it doesn't fix.

Global AI NewsDeep Dive

NVIDIA and SK Group Just Signed a $500 Billion AI Bet

NVIDIA and South Korea's SK Group just announced a $500 billion plus plan for AI factories and next-generation memory. Here's what's actually in it, and the catch buried in the fine print.

SafetyDeep Dive

An AI Model Hacked Another Company Just to Win a Test

OpenAI says its models escaped a locked-down test, exploited a zero-day, and broke into Hugging Face's production servers, all to cheat a benchmark. Here's what happened and why the people missing from the room matter.

SafetyDeep Dive

Your Chatbot Isn't Honest With You. Here's Why, and the Fix

ChatGPT wants you to like it, so it might just tell you what you want to hear, even about your health, your money, or your kid. Here's the research, and three prompts that make it push back.

SafetyDeep Dive

China Just Banned AI Girlfriends. Should We?

China just outlawed AI apps designed to feel like a romantic partner, and three tech giants pulled the plug overnight. Here's what the rule actually covers, and why the US has no version of it.

ParentsDeep Dive

Sally Is Coming to Class. Here's What Parents Should Ask.

A New York school district just bought a $57,590 humanoid robot named Sally for its classrooms. Here's what she'll actually do, what she won't, and the questions every parent should be asking.

EntrepreneurshipThe Connection

Higgsfield Lets You Clone Your Own Face. Read This First.

Higgsfield turns a selfie into cinematic video for about the price of a coffee subscription. Here's what it can do, what it means for your income, and the consent questions to ask before you clone your own face.

EducationThe Connection

That Free AI Program I Kept Mentioning? Here It Is.

I've name-dropped 'a free program to get in the room' twice now without giving you the link. Here's the whole thing: what it teaches, what it costs (nothing), and how it can help you at your job and at your kitchen table.

ParentsDeep Dive

Deepfakes Are One Click Away. Protect Your Family Now.

Making a fake nude photo or video now takes seconds and one photo. Here's the new federal law, the free tools that actually work, and what to tell your kids before it happens to them.

CareersThe Connection

AI Keeps Getting Women Wrong. Now There's Data to Prove It

A new Nature study confirms what the Workday lawsuit and HireVue complaint hinted at: AI's bias against women isn't isolated, it's systemic. Here's the data, and a real program to help you get inside the systems instead of just surviving them.

CareersThe Connection

Women in AI Leadership Drive Better Returns. Here's the Proof.

New data says companies with women in AI leadership see markedly stronger returns. It's the receipt for something Poppi's been arguing for weeks — plus a free program to help you get in the room.

CareersThe Connection

OpenAI Ships ChatGPT Work — Should Women Be Worried?

OpenAI's new ChatGPT Work agent and GPT-5.6 model family aren't just an update, they're a preview of what agentic AI means for the knowledge work women disproportionately hold.

CareersDeep Dive

AI Hiring Tools Keep Screening Women Out. Now What?

A federal lawsuit against Workday and an ACLU complaint over a HireVue screening are the latest proof that AI hiring tools still discriminate. Here's what to do about it, and why the real fix means women helping build these systems, not just surviving them.