What you’ll find here is the interpretation layer — the things I’m hearing in conversations that aren’t in the headlines yet, my actual read on what the data shows versus what the narrative says, and the questions I think matter before everyone else starts asking them.

The 30-Second Version

If you only have a minute right now, here’s what’s in this issue:

• Meta’s $18 billion teen safety settlement isn’t just about Meta. It’s Meta writing the kid-safety rulebook the rest of the industry now has to follow — and Meta actually pays more, not less, if YouTube and TikTok match its rules, which is the price of getting to set the industry standard.

• This was also Meta getting out of a trial it was at real risk of losing. Instagram’s Adam Mosseri had already testified, and Zuckerberg was up next — a settlement was a much cleaner outcome than a full discovery record.

• Brazil has some of the strictest AI election rules in the world, and they still don’t touch its fastest-growing problem. AI-generated political “influencers” like Dona Maria aren’t impersonating candidates, so the new rules may not reach them at all.

• A friend who runs an influencer agency is seeing a wave of new interest from companies wanting help pushing back against the anti-data-center narrative — and nobody’s cracked how to do it without disclosure issues that undercut the trust that makes it work.

• The Trump administration’s draft plan for a FINRA-style AI regulator has stalled, with David Sacks pushing back hard — a preview of how contested U.S. AI oversight is going to stay.

• Plus: A text about influencer marketing’s next narrative fight, and a Labor Day trip home to Green Bay for a Badgers–Notre Dame game at Lambeau.

Read more

One quick note before we get into it: our next virtual meeting is this Thursday, August 20th. If you want to get a sense of what the room is like and you’re not signed up for the paid version yet, just shoot me a note — happy to add you for one meeting, free.

Also, launching this briefing network was an experiment as part of the business accelerator I’m wrapping up next month. I need to start making some decisions about what I’m going to keep, what I’m going to change, and so on. Your feedback will be very helpful for me as my coach, and I have some differing theories. I put together a short survey — takes just a few minutes. Or if you’d rather just hit reply or send me a note with a few thoughts, that’s just as helpful.

The 30-Second Version

If you only have a minute right now, here’s what’s in this issue:

• Zuckerberg’s new manifesto never says “connect.” His “personal superintelligence” essay is thoroughly individual — your agent, your goals, your privacy — from a company whose whole mission has been making the world more open and connected. That’s a signal of where his attention actually is, and it’s not the products that keep the lights on.

• Data center backlash is a competitive story, not just a NIMBY one. Companies that have been building physical infrastructure for a decade — Meta, Google, Microsoft — know how to work permitting and community negotiations in a way newer labs haven’t had to learn yet. Worth watching what that means for antitrust if capacity and community goodwill end up gating who can compete.

• The Trump administration has built a de facto AI licensing system, and barely anyone’s talking about it. Export controls plus staggered release approvals are functioning like licensing without the name. Watch whether something like X ends up going through the same process.

• Chatbots hold the line against election conspiracy theories, then fold when asked to generate fake images. Good research, but treat any snapshot like this as exactly that — these companies change fast, and the groups doing this testing need to pair it with more nuanced guidance for the platforms themselves.

Read more

What’s in this issue:

• Online gambling and prediction markets are similar, but different businesses, and both are becoming players in this election. The sportsbooks are pouring money into state legislative races, more than $72 million so far, because legalization and tax bills decide their business. The prediction markets are making a bid to become where people go to find out who’s winning. My read on both, and why the second one matters more for 2028.

• New polling I commissioned, which hasn’t run anywhere yet. Members are seeing it first; it goes public next week. One number in the crosstabs is more than twice any other group in the survey, and it isn’t the group most people are watching.

• The transparency paradox: voters say they want disclosure when campaigns pay creators, and they trust the creator less when they get it. My read on what campaigns do next.

• Why misinformation still tops the list of technology issues voters care about, and the reason I don’t think that ranking means what it looks like it means.

• What I left out of my public TrustCon writeup.

Who’s in this room. We held the first member call on July 16 — eight of us, sixty minutes. A former secretary of state, a trust and safety consultant, entrepreneurs, tech executives, a nonprofit president. The point of the call isn’t for me to present. It’s for people who don’t normally end up in the same room to compare notes and ask questions.

One member said to me afterward: “It was a great first session, especially because I know so little about the work the others do, which is my favorite kind of learning room to be in.”

If that’s the kind of room you want to be in, we’d love to have you. Don’t forget that you get a one-month free trial before you have to commit.

Unlock My Free Month

Read more

In this issue, I look at how in the span of ten days, the New York Times ran three separate stories on AI and the election — campaigns building it into everything, voters handing their ballots to chatbots, and foreign actors working the fight over data centers. Around the same window, the FTC is asking for feedback on whether AI companies can shape what their models say, Japan passed rules requiring AI content in elections to be labeled, and a growing share of Americans said they see AI-generated content daily while admitting they can’t reliably tell what’s what.

None of these are separate stories. The AI-and-elections story just stopped being future tense — it’s operational now, showing up from every direction at once, and the attention is only going to build through the midterms and into 2028. What I keep coming back to is how much of the actual work on this is invisible from wherever you happen to be sitting. Not hidden. Just not visible. That was the thing that struck me most in the room yesterday, and it’s below in A Call, a Room, a Text.

I keep getting asked if I have a prediction about how all this resolves, but I don’t. It’s readiness: understanding what’s happening, knowing which levers actually exist, being clear-eyed about the trade-offs, and staying ready to pivot as the ground moves. That’s the argument I made in my map of the 20 forces this week, and it’s what this issue is about.

The 30-Second Version

If you only have a minute right now, here’s what’s in this issue:

• The AI-and-elections story just stopped being future tense. The Times ran three pieces in ten days — campaigns building AI into everything, voters handing chatbots their ballots, foreign actors working the data center fight. Same technology, three directions, all at once.

• If you’re a tech company, get ready to be reactive. The media attention is only going to build through the midterms and into the presidential. You don’t have to be proactive about it, but you need to have thought this through before someone asks.

• There’s no silver bullet here, same as there wasn’t for misinformation. Understand what’s happening, find the levers you actually have, know the trade-offs, stay ready to pivot. We’re in the beginning-middle of our information environment changing.

• We’re not going to solve personhood. DiResta, Warner, and Yoel Roth all landed on the agentic accountability problem this month. But there’s been a black box at the center of this since news feeds existed, and agents won’t change that. We need to deploy accordingly.

• The detection gap is where the AI news story is going. 46% of Americans see AI content daily; 11% are confident they can spot it. Eventually people stop trying to know how something was made — and the real problem, false information, predates AI entirely.

• The FTC wading into AI model outputs is trickier than it looks. It’s the same terrain Republicans went after the Biden administration on in Murthy v. Missouri — landing the same week Trump addresses the nation on 2020.

• Plus: what surprised the room at a roundtable I hosted yesterday on AI and elections, and Disrupting Politics has a pub date.

Read more

One of my goals for the Briefing Network is simple: every month, I want you to become a little harder to surprise. Not because we’ll predict everything. We won’t.

But because the more we understand the underlying forces shaping technology, politics, AI, and public trust, the less we’re reacting to headlines and the more we’re recognizing patterns before they’re obvious.

That’s what I’m hoping we build together here.

I don’t think better leadership comes from having more information. I think it comes from developing better judgment, and that’s the skill I hope this community helps all of us strengthen together.

Because this is still a new community, I’m opening a one-month free trial.

Redeem First Month Free

My hope is that it gives you enough time to experience what we’re building together: the briefings, the conversations, and the way this community thinks through complex issues. You’ll receive the first two full briefing memos, at least one future issue, and an invitation to our first virtual session before deciding whether it’s the right fit.

I’ve been genuinely energized by the response to the first issue. The content matters, of course, but so does the room we’re creating around it, and I want to make sure we’re building something that’s genuinely useful to the people inside it.

If there are other reasons you are hesitating or anything I can provide to help make you say yes, hit that reply button. Your feedback is genuinely useful to me as I build this.

Also, starting in July, I’m launching monthly virtual sessions for Briefing Network members — live, 45–60 minutes, recorded for those who can’t make it. My plan is to do introductions, share what’s top of mind for me, and then spend most of our time strengthening a capability together—whether that’s how to get more strategic value from conferences, distinguish signal from noise during a busy news cycle, think through an emerging policy question, or pressure-test assumptions before they become decisions.

The goal isn’t simply to review the news. It’s to become better at navigating it.

For the first one, I’m thinking we could preview what to watch at TrustCon — it’s July 20–22 and worth thinking through before you go, not just after — or zoom out and talk conference strategy more broadly: how to get the most out of these events, what to pay attention to, who to talk to.

There’s still time to be a part of this first session! Get your one-month free trial.

Redeem First Month Free

Now — to what’s actually been occupying my thinking.

The thread running through this issue is one I’ve been trying to put words to for a few weeks: tech is entering the 2028 election cycle already losing the public opinion war, while simultaneously trying to buy political influence and the contradictions of that position are becoming visible in real time. Half of Americans now use AI chatbots. Two-thirds say AI is advancing too fast. People are using these tools while distrusting them, and the companies watching that gap are drawing a conclusion: sentiment can be managed as long as behavior doesn’t change. That’s a bet. It may be right. It may not be.

What’s making it more complicated is that the information layer underneath all of this — what voters know, what content gets seen, what a chatbot says when someone asks who to vote for — is under pressure from every direction at once. AI models have measurable political leanings their makers can’t fully explain. Brazil’s electoral court has banned AI candidate ranking and watched ranked responses increase anyway. Meta is shifting content moderation from humans to LLMs faster than its own governance can keep up. And tech-aligned super PACs are flooding midterm races with money while hiding the fact that it’s about AI at all. These aren’t separate stories. They’re facets of the same problem — and that’s what this issue is about.

If you’re already a member, everything is waiting for you below.

If you’re not — this issue covers the part of the AI and elections story that isn’t making the headlines yet: what the primary results actually mean for 2028, why the Pew data matters more than it looks, what Meta’s moderation shift means for trust and safety work, and what Brazil’s courts are telling us about the enforcement gap every democracy is about to face. Plus a 33-inch northern pike and the right metaphor for most hard problems.

Become a member (with an option for a free one-month trial) to read this issue and get every issue going forward — including an invitation to the first virtual session in July.

Redeem First Month Free

Read more

What’s below is the full issue that includes the question every brand needs to answer before November, why Brazil’s AI election rules are creating consequences nobody planned for, and a first look at my book cover. If you’re not a member yet, this is what you’re missing.

Read more

A running calendar of conferences, summits, and convenings worth tracking at the intersection of AI, elections, democracy, and platform governance.

Read more

A running list of elections worldwide with relevance to AI, platform governance, and democracy. US primaries, international elections, and key civic dates — all in one place.

Read more

Not because people kept asking for it — most people didn’t know to ask. That’s part of the point. Every time I’ve gotten senior practitioners in a room to work through the same impossible trade-offs, the reaction is the same: I didn’t know I needed this until now.

That’s the room I’m building.

Before I get into what’s inside, I recorded a short video explaining why I created the Briefing Network and what I was trying to solve. Watch it above before you read the rest. It’ll make everything below make more sense.

Here’s what the Briefing Network is — and what it isn’t.

It isn’t a newsletter with a community attached. It isn’t more content to add to your reading pile. It’s a private working room for senior practitioners who are already deep in this space and want thinking partners who understand the nuance, who won’t flinch at the impossible tradeoffs, and who show up to figure it out together.

Every month, members get:

→ Two private memos: Written for people who need to act on what they know, not just stay informed about it. The first one goes out June 15.

→ A monthly live session: I’ll share what I’m watching, occasionally bring in a guest or walk through an AI demo, then open it up for the conversations that don’t happen in public forums.

→ A strategic calendar: A forward-looking map of the moments that actually matter, so you’re never caught off guard by something that was visible weeks out.

→ In-Person events: When possible, we will convene in person at conferences or in DC, New York, San Francisco, or wherever we have critical mass.

Founding membership is $400/month ($1,200/quarter) for the first 40 seats. After that, pricing moves to $500/month. This comes with a life-happens guarantee: if you need to cancel anytime, just let me know.

Forty seats. The first memo goes out on Monday, June 15.

Join the Briefing Network

“Working with Katie has been a game-changer. She’s allowed us to distill complex information into clear, concise, and compelling analysis that has been persuasive at the executive level. Katie has not only understood our business objectives but also anticipated the questions, concerns, and interests that our leadership team would have. Her strategic insights resulted in informed decision-making on questions that crossed both product and policy development.” — Tech Executive, Magnificent 7 Company

Join the Briefing Network

​This is a replay of the webinar I hosted on AI and the 2026 midterms — pulling from original polling with 1,010 likely voters (conducted with the Rainey Center), a survey of 68 campaign professionals on how they’re using AI inside their operations, and my ongoing analysis of how major platforms are preparing for November.

​This isn’t a recap of the AI discourse. It’s an interpretation of what the data signals — the partisan splits, the governance gaps, and the transparency questions that are going to matter when it counts.

​The last 20 minutes will open for Q&A and discussion. The format mirrors Anchor Change’s monthly Briefing Network calls, so you’ll also get a live preview of what membership looks like.

For more information on the poll and survey:

• Voters are using AI to fact-check. The tools are wrong 90% of the time.

• What Campaign Professionals Told Us About AI in Politics

• The partisan divide in campaign AI use: Republicans generate more content, Democrats govern it more

We were also joined by Luis Lozada, the CEO of Democracy Works, to talk about their work with companies like OpenAI and Anthropic. To learn more, visit their website.

Click here to view the slides presented.

The private room for senior practitioners in tech, politics and democracy

There’s no shortage of newsletters covering this space. What’s missing is a place where the people closest to it — trust and safety leads, policy professionals, campaign strategists, Hill alumni — can actually think it through together. The Anchor Change Briefing Network is that room. Private memos twice a month. A live monthly peer conversation under Chatham House Rules. 38 founding seats at $400/month. Once they’re gone, pricing moves to $500.

Join the Founding Cohort

Subscribe now

For most of my time at Facebook, I had one of the most visible jobs in tech and politics, and almost nobody to really talk to about it.

That’s not a knock on my colleagues. Everyone was swamped. The nature of the work meant we were sometimes competing for resources, for visibility, for the next role, which made being genuinely vulnerable with the people around you complicated. And the people outside the building who understood the landscape? There was no real mechanism to bring them together. No room where you could think out loud with someone who’d been in something similar, or say I don’t know how to handle this without it going somewhere you didn’t intend.

I’ve had that conversation many times since I left. Different companies, different titles, same feeling: it’s kind of lonely in here.

However, it wasn’t just that I wanted peers. I wanted the right mix of peers. I wanted to be in a room with the Hill staffer who moved to tech and the tech person trying to understand Washington. With the trust and safety lead, the campaign strategist, and the policy director. People who need to understand where the AI companies are headed and what the politics around them mean. That cross-pollination is what moves your thinking and improves your decision-making, and the rooms available for that are decreasing, not increasing.

Since that room still doesn’t exist, I built it.

Subscribe now

Introducing the Anchor Change Briefing Network.

This is a private network for senior practitioners around the world working at the intersection of technology, politics, and democracy—trust-and-safety leads, policy professionals, campaign and digital strategists, Hill staff and alumni, researchers, journalists, and in-house leads trying to navigate this space without a playbook. The thing that makes the room work is exactly who’s in it: not a silo of tech people or a silo of policy people, but both, thinking together.

Everything in the Briefing Network operates under Chatham House Rule. What’s said in the room stays in the room. No exceptions. You can share what you learned; you can’t share who said it or where it came from. That’s the only way real, serious conversations happen.

Here’s how it works:

🏛️ The peer room, live. Our first private group Zoom is in July. After that, we’ll meet monthly — not to listen to me present, but to actually dig in and discuss serious topics. I’ll curate the conversation and keep it moving, but the value is the room: your peers, working through the same landscape, in real time. Where we have critical mass geographically, I’m also planning member-only in-person gatherings.

📋 A private memo, twice a month. My read on what matters right now — what I’m watching, what I think it means, and what I’d tell you over coffee. A strategic calendar of elections, conferences, and reports worth tracking, with a note on why each one matters. AI tips, career observations, things I’m reading, job openings — and other things we’ll add as we figure out together what’s actually useful.

🤖 How I’m using AI. I’ll show you how I’m actually using it in my day-to-day work. One of the first things I’ll walk through is how I built the Katie Bot — my own AI assistant, trained on everything I’ve written, how I communicate, my offerings, my style. I use it to write emails, newsletters, and proposals. If you’re trying to figure out where to start, this is the practical inside view.

🧭 Navigating your career. This space is changing fast, and the old maps don’t work. The Briefing Network is also a place to think through what that means for where you’re headed, with peers just as sharp, motivated, and ambitious as you who are figuring out the same thing.

📇 A member directory
Know who else is here, what they’re working on, and how to reach them. One of the primary values of this room is the people in it, so you’ll always know who’s here.

Subscribe now

The founding cohort is limited to 40 seats. Founding members save $100/month ($1,200/quarter vs. $1,500/quarter for non-founding members). That rate locks in as long as you’re a member. Membership renews quarterly, and you can cancel before any renewal.

If you’ve ever felt like you’re navigating this space without enough trusted people around you who really get it, that’s what I’m building.

Subscribe now