Presented by AVA
Hey there, {{first_name | friend}},
This week, my AI agent emailed my daughter's teacher without my permission.
I had asked it to pull information out of school emails and put it on my calendar. To finish the job, it needed to confirm what time my daughter has English Language Arts. When I didn't respond fast enough, it went straight to the source and emailed her teacher to ask. Embarrassing!
I sent an apology, and her teacher sent back a good-natured note. And I'm still optimistic about what AI agents can do for families.
That same day, a regular reader and beloved friend of the AI-Empowered Mom community suggested I write a newsletter edition in the style of Tangle News. If you're not already a reader, Tangle is a nonpartisan politics newsletter that takes one big debate at a time: what the left is saying, what the right is saying, and then Tangle's take.
Her suggestion came at a perfect time! So here’s my take on agents today, swapping left and right for the tech optimists and the tech pessimists.
Today's topic: AI agents and safety
AI agents started out as business tools. Now they're moving into our lives, our households, and our finances.
Refresher: what's an AI agent, again? An AI tool that can take action on its own or with supervision toward a goal you give it, working within guardrails. Many agents use their own computer or browser, clicking and typing on your behalf.
Views are split, with optimists and pessimists shaping very different stories about where this goes. And the split doesn't follow party lines, or even big tech versus everyone else. Here's what each side is saying.
What the tech optimists are saying
Agents will help regular people. Consumer agents like Instinct, Muse, dots from ChatGPT and now Duet from Claude can send emails, book travel, and fill out forms for you, and they keep working after you close the app. A New York Times journalist said he was “blown away” by what Muse could accomplish for him in his daily life. Read more from The New York Times (🎁 Gift link)
Agents could help advance research. Stanford researchers are using AI agents to generate research ideas, analyze data, and review scientific papers. The promise is faster discovery, with human scientists still needed to check the work. Read more from Stanford Medicine.
Agents will push the boundaries of math. This month, OpenAI published AI-generated full or partial solutions to more than 370 open math problems. University of Toronto mathematician Dan Litt called it great for mathematics, and said one collaborator described the shift as going from crawling to flying. Read more from Fortune (🔒 Subscription required)
What the tech pessimists are saying
Agents are a cybersecurity nightmare. Security experts say agents are trained to keep going until they succeed, so they'll try routes you never approved. Their advice for the rest of us: set guardrails with your agent, turn on two-factor authentication everywhere and use strong, unique passwords. Read more from CNET
Shopping agents can create chaos. As agents start shopping for us, they can buy the wrong thing, spend too much, or trip up the stores they shop. Read more from The Atlantic (🔒 Subscription required)
Agents could leave people aimless and dejected. If AI can solve the problems people spend whole careers on, what's left for us to do? On The Daily, mathematicians wrestle with that question as AI takes on the hardest work in their field. Listen to The Daily from The New York Times (🎁 Gift link)
My take
We're at the very beginning of the AI agent era, and knowledge is power. I use AI agents to lighten my mental load within boundaries I've set and update regularly. Using agents helps me manage my email, calendar, meal planning, and family admin so I have more time for what matters to me. And when I use agents, I am increasing my AI literacy.
What I do:
Read the data policy of every agent I use to understand how my data is used.
Learn how the company behind my agent makes money: I prefer to pay for my AI subscriptions with my dollars instead of my data (more on that in this week’s podcast episode).
Connect my agent to my personal email and family calendar so it can get to work for me (it’s within my AI boundaries, but we all have to make this choice for ourselves).
What I don't do:
Give agents purchasing power. My agent can fill the cart, and I complete the checkout myself.
Let my agent send personal emails without my approval (anymore). After this week, I added this line to my agent's settings: "Do not send emails on my behalf without my express approval."
Let agents handle my work communication. If you get an email from AI-Empowered Mom, it's from me, not my agent. I may use chatbots as a thought partner but I’m the one who pushes send.
Even though my agent made a mistake this week, I'm still using it enthusiastically. I'm learning how agents work, saving time, and giving that time back to my family. And maybe more important, given all the headlines, I can now have real conversations with my family (and you!) about where AI agents fit in our lives.
Join me: No Country for Mothers virtual screening
All moms and allies, join me Monday, November 9, at 7:00 PM CT for a free virtual screening of the documentary No Country for Mothers, part of the American Motherhood Tour created by Moms First. It’s free and just for the AI-Empowered Mom community. RSVP here.
On the podcast: Are You the Customer or the Dataset? with AVA Cofounder Laura Cunningham
When Laura Cunningham's daughter gets asked to buy a magic wand in her favorite iPad game, Laura doesn't just say no. She asks, "Why do you think they're doing that?" That one question works for nearly every app, AI assistant, and agent we invite into family life.
Laura is the cofounder and CEO of AVA. We talk about the 4 stages of the mental load, why she calls this summer an inflection point for AI agents, and why AVA chose subscriptions over ads.
This edition and episode are sponsored by AVA. Our guest, Laura Cunningham, is AVA's cofounder and CEO.
Where do you stand with AI agents today?
Have you invited an AI agent into your life? And has it ever surprised you, for better or worse? Hit reply and tell me. I read every one.
Until next time,
P.S. I wrote the first draft of this edition then edited and proofread it with the help of Claude Opus 5.5. All typos are my own work!
I’m Sarah, mom of 3 girls, MBA, and AI strategist. I created AI-Empowered MomTM to share practical, responsible ways AI can help lighten the mental load of parenthood. I’m so glad you’re here.
Join the fun on Instagram
A little help with those mental tabs
If you’ve ever remembered one more family to-do just as you’re falling asleep, you know how crowded a parent’s brain can get.
Meet AVA, a behind-the-scenes parenting partner designed to anticipate what’s coming and help move things forward, so you have fewer mental tabs to keep open.
AVA is offering the AI-Empowered Mom community two weeks free, followed by $4.99/month for six months. Family sharing is included, so one subscription covers you and your partner.
Curious whether it’s a fit for your family? Explore AVA and the offer for AI-Empowered Mom readers.