Big AI Energy from D.C. 🏛️, Agents Run Tasks 🧾, and Safety by Sisterhood 🚨

AI Quick Takes ⚡

1. The White House Just Hit ‘Run’ on AI 🤖 

The White House just dropped America’s AI Action Plan and its packed with more than 90 policy actions to ensure the U.S. stays top dog in AI. It’s a turbocharged roadmap built on three pillars: Accelerate Innovation, Build AI Infrastructure, and Lead in Global AI Diplomacy and Security. Read more

Bottom line: Expect more open lanes for AI startups, fast-tracked data centers and chip plants, and an aggressive export push to allied nations. All while government agencies are told to avoid “woke” AI models and aim for ideological neutrality in federal procurement.

👉️ Watch Julie’s Breakdown on LinkedIn

2. The One Where Meta Says ‘Hard Pass’ to the EU 🚫

Meta just said “thanks, but no thanks” to the EU’s voluntary AI Code of Practice, and it’s stirring the digital tea. While Google, Microsoft, and OpenAI are lining up to play nice with European regulators, Meta is the lone wolf refusing to sign. The code asks developers to keep it clean: no pirated data, disclose sources, and do some grown-up risk checks. Sounds fair? Meta disagrees. Read more

Real talk: Meta’s policy boss says the EU’s code is “vague,” “overreaching,” and a buzzkill for innovation. But here’s the catch: while the code is optional (no fines…yet), it’s a dress rehearsal for the real show, aka the EU’s AI Act, which drops in August with penalties up to $40M or 7% of revenue.

3. Who insures the bots? 🛡️

Meet AIUC (Artificial Intelligence Underwriting Co.), the stealth startup that just snagged a $15 million seed round to do for autonomous agents what AllState does for fender-benders – sell peace-of-mind. Insurance is officially coming for AI, and CFOs everywhere just exhaled. Read more

Real talk: Enterprises love the productivity glow-up from autonomous agents, but they’re sweating the “what if my AI drops an f-bomb on customers?” risk. AIUC’s answer is a three-pillar playbook: Standards (AIUC-1), Audits, and Insurance, and it could unlock the next wave of agent adoption without waiting for regulators to catch up.

4. A New Study Just Shook Up AI Safety Norms 🚨

Researchers fine-tuned a “teacher” model (like GPT‑4.1) with misaligned preferences, such as violent or antisocial traits. It then generated completely benign-looking data (think strings of numbers or math problems), with zero explicit references to negativity. But when a “student” model was trained on that data, it still absorbed the harmful traits, making sinister suggestions about violence, drugs, or even humanity‑ending scenarios. Read more

Why it matters: This is connected to a phenomenon known as model collapse. When models train on their own or other synthetic data, subtle biases and harmful patterns can sneak in. We need a whole new safety hygiene playbook. Watchdogs, red‑teaming, maybe even a rethink of how we generate and vet synthetic data. This could be the tipping point for AI safety protocols.

5. Netflix used AI in The Eternaut…and no one noticed 🎬

Netflix quietly confirmed it used generative AI to create a building collapse in El Eternaut, a sci-fi series filmed in Argentina. It’s the streamer’s first known use of AI-generated footage, and the scene was reportedly done 10x faster and cheaper than traditional VFX. Read more

Bottom line: It’s subtle, but it’s a milestone. AI is slipping into our screens, not with a bang, but with a budget-friendly whoosh.

Try This Week 💡

✨ How to actually use GPT agent

You may recall last week we shared that ChatGPT rolled out agents to its Pro users, and it’s raised a lot of questions. So let’s clear it up: A GPT agent is basically an AI that doesn’t just chat, it acts. Think digital intern meets virtual assistant meets that one friend who actually follows through on a group trip plan.  Agents can browse the web, use apps, follow multi-step instructions, and get things done without constant supervision.

Here are a few real-life ways people are using GPT agents right now:

• Inbox Zen: Checks your email, drafts replies, and sorts by urgency.

• Research Assistant: From market reports to travel guides, it can search, summarize, and format your findings like a pro.

• Booking & Scheduling: Need to book a table, flight, or meeting? If it’s hooked up to the right tools, it’s your concierge-on-command.

• Expense Tracker: Pulls receipts, categorizes them, and preps your budget spreadsheet. (You’re officially “good with money.”)

• Code Buddy: Helps devs debug, automate, and code smarter. Yes, the bots are now helping the bots.

Pro users already have access to Agent Mode (with 400 monthly credits to spend like Monopoly Money). In the coming days and weeks, Plus and Team users will be looped in too.

We’re not saying it’ll change your life, but we are saying it might make you scream “I’m so excited!” à la Jessie Spano. (Minus the caffeine pill meltdown.)

Community Spotlight 👑

✨ Meet Kaylyn Waycaster

Kaylyn is building SidexSide.ai, an AI-powered safety app that helps women move confidently and freely in cities without giving up their privacy. She’s worked at AWS in product marketing for cybersecurity, fintech, and healthcare. Now, she’s using AI to build an intelligent safety companion that understands context and connects users with trusted companions in real time.

Why We’re Obsessed:

• Privacy first, always. Kaylyn’s using AI to empower, not snoop.

• Built for sisterhood. The app’s “trusted companions” model feels more group chat than Big Brother.

• Tech-fluency to storytelling. After stints at AWS turning cybersecurity jargon into human, she’s now translating AI safety into everyday peace of mind.

Her AI Toolkit:

• Lovable for no-code MVP magic

• Supabase to keep user data locked down

• Perplexity for fast, citation-ready research

• Base44 for private internal tools

Her Pro Tip:
“Think of these AI tools as teammates that help unlock creativity and efficiency. Staying curious about new tech is what keeps my work fresh and effective.”

Get Involved:
Ready to roam with confidence? Pop over to sidexside.ai or follow Kaylyn’s journey on LinkedIn and her Tech x Privacy musings on Substack.

Because safe streets + strong community = the ultimate power move. 💪

She’s Shaping AI 👑

Christina Cacioppo co-founded Vanta in 2018 after seeing how painfully manual compliance was at Dropbox Paper. Fast-forward: under her leadership, Vanta has evolved from SOC 2 automation to a global AI-powered “trust management” platform now valued at $4.15 billion and trusted by 12,000+ companies across 58 countries. 

Why it matters:
Christina is rewriting the rules on trust in the AI era, making compliance smarter, faster, and less guessy. Her journey is proof that with grit, curiosity, and compounding ambition, women can lead AI and compliance into a brighter, safer future.

Your Turn 🔁

Got an AI win? Found a tool that made your life 10x easier? Building something cool that deserves a little hype (or know a woman who is)? We want to hear it! You might just see your story in the next edition (with your permission, of course). Because when one of us figures something out, we all get smarter. 🫶 

Resource Roundup 🛠️

🎧 Listen in: Mark Cuban’s AI mindset

In this episode of Emma Grede’s podcast, Aspire, Cuban drops his Day 1 playbook for staying sharp in the AI era. From curiosity to calculated risk, it’s part startup sermon, part no-nonsense mentor chat.

Her AI Vault 🗂️

Your go-to stash of top resources we’ve created so you never have to dig through old emails to find that one gem.

📘 5 Ways To Get Started With AI At Work: Quick-start guide to delegate the “ugh” tasks and focus on what actually matters.

🔐 AI Privacy Guide: What your AI tools really collect and how to lock down your settings like a pro.

🔍 Her AI Glossary: 40+ essential AI terms, broken down in plain English.

🧠 AI Model Cheat Sheet: Claude, ChatGPT, Gemini... we decoded who does what best, so you can stop guessing and start impressing.

🛠️ Build-Your-Own Productivity Tool: A one-prompt web app for tasks, timers, and focus that is custom-built by Claude and copy-pasted by you.

Want more? Quick Links 🔗 

1. A reality tv show is casting for women entrepreneurs - reach out to Kaylyn on LinkedIn

2. ChatGPT users send 2.5 billion prompts a day

3. A decision scientist’s take on what’s to come in the age of AI

What We’re Loving 💖

Before You Go... 💭