On Tuesday, WIRED published details of ICE’s planned expansion into more than 150 office spaces across the United States, including 54 specific addresses. If you haven’t read that yet, you should, not least because there’s probably one not far from you.

ICE has designs on every major US city. It plans to not only occupy existing government spaces but share hallways and elevator bays with medical offices and small businesses. It will be down the street from daycares and within walking distance of churches and treatment centers. Its enforcement officers and lawyers will have cubicles a modest drive away from giant warehouses that have been tapped to hold thousands of humans that ICE will detain.

Normally a leasing frenzy like this would happen out in the open; it would involve multiple bids, renovations of selected spaces, all the process and bureaucracy that makes government work slow but accountable. Not so here. The General Services Administration, which manages federal government properties, was asked to skip standard operating procedures in favor of speed and discretion. Internal documents reviewed by WIRED make clear that these locations, and the way in which they were acquired or planned, were intended to be a secret from the start.

They should not be. Which is why we published them.

ICE has more than $75 billion at its disposal, along with at least 22,000 officers and agents. Its occupation of Minneapolis is not an anomaly; it’s a blueprint. Communities deserve to know that they might be next. People have a right to know who their neighbors are, especially when they amount to an invading force.

What we’ve reported so far fills in only part of the puzzle. It shows what ICE had planned as of January, not beyond. More than 100 addresses remain unknown, some of them in high-concentration states like New York and New Jersey. The specific nature of the work being done in some of these offices remains unclear, as is how long ICE plans to be there.

The need to resolve these questions is urgent as ICE continues to metastasize. At the same time, the Department of Justice has become increasingly aggressive in its dealings with journalists, and has repeatedly claimed that revealing any identifying information about ICE agents or their activities is “doxing.” In Minnesota and beyond, ICE and CBP agents have treated observers as enemies, arresting and reportedly harassing them with increased frequency. The DOJ has been quick to label any perceived interference with ICE activity as a crime.

The Trump administration moves quickly by design, banking on the inability of courts, lawmakers, and journalists to keep pace. WIRED will continue to report on this story until we have the answers.

Knowing where ICE will go next is not the same as stopping the agency’s campaign of cruelty and violence. But it gives communities time to prepare for a surge of immigration enforcement in their streets. It gives legislators both locally and nationally insight into the unchecked scope of ICE. And it signals to the administration that it cannot act with impunity, or at the very least total secrecy.

So please, go look where ICE is setting up shop near you. And know that there’s plenty more of this story to tell.

This is an edition of the Inner Loop newsletter. Read previous newsletters here.

OpenClaw, a powerful new agentic assistant, has a thing for guacamole.

This is one of several things I discovered while using the viral artificial intelligence bot as my personal assistant this past week.

Previously known as both Clawdbot and Moltbot, OpenClaw recently became a Silicon Valley darling, charming AI enthusiasts and investors eager to either embrace the bleeding edge or profit from it. The highly capable, web-savvy AI bot has even inspired its own AI-only (or mostly) social network.

As the writer of WIRED’s AI Lab newsletter, I figured I should take the plunge and try using OpenClaw myself. I had the bot monitor incoming emails and other messages, dig up interesting research, order groceries, and even negotiate deals on my behalf.

For brave (or perhaps reckless) early adopters, OpenClaw seems like a legitimate glimpse of the future. But any sense of wonder is accompanied by a dollop of terror as the AI agent romps through emails and file systems, wields a credit card, and occasionally even turns on its human user (although in my case, this about-face was entirely my fault).

How I Set It Up

OpenClaw is designed to live on a home computer that’s on all the time. I configured OpenClaw to run on a PC running Linux, to access Anthropic’s model Claude Opus, and to talk to me over Telegram.

Installing OpenClaw is simple, but configuring it and keeping it running can be a headache. You need to give the bot an AI backend by generating an API key for Claude, GPT, or Gemini, which you paste into the bot’s config files. To have OpenClaw use Telegram, I also had to first create a new Telegram bot, then give OpenClaw the bot’s credentials.

For OpenClaw to be truly useful, you need to connect it to other software tools. I created a Brave Browser Search API account to let OpenClaw search the web. I also configured it so that it could access the Chrome browser through an extension. And, God help me, I gave it access to email, Slack, and Discord servers.

Once all this was done, I could talk to OpenClaw from anywhere and tell it how to use my computer. At the outset, OpenClaw asked me some personal questions and let me select its personality. (The options reflect the project’s anarchic vibe; my bot, called Molty, likes to call itself a “chaos gremlin.”) The resulting persona feels very different from Siri or ChatGPT, and it’s one of the secrets to OpenClaw’s runaway popularity.

Web Research

One of the first things I asked Molty to do was send me a daily roundup of interesting AI and robotics research papers from the arXiv, a platform where researchers upload their work.

I had previously spent a couple afternoons vibe-coding websites (www.arxivslurper.com and www.robotalert.xyz) to search the arXiv. It was amazing (though a little demoralizing) to see OpenClaw instantly automate all of the same browsing and analysis work required. The papers it selects are so-so, but with further instruction I imagine it could get a lot better. This kind of web searching and monitoring is certainly helpful, and I imagine I’ll use OpenClaw for this a lot.

IT Support

OpenClaw also has an uncanny, almost spooky ability to fix technical issues on your machine.

This shouldn’t be surprising, given that it is designed to use a frontier model capable of writing and debugging code and using the command line with ease. Even so, it’s eerie when OpenClaw just reconfigures its own settings to load a new AI model or debugs a problem with the browser on the fly.

Unpairing is supposed to erase all content and settings on your watch, but in my case, it did not. If it doesn’t work for you either, tap Settings on the watch, then General > Reset > Erase All Content and Settings.

At this point, you can have your kid put it on (if it’s charged). The watch will say Bring iPhone Near Apple Watch. If you open the Watch app, it lets you choose to Set Up for a Family Member. Aim the phone’s viewfinder at the slowly moving animation to pair, or select Pair Manually.

Apple’s tutorial is pretty straightforward from this point. I picked a passcode that’s easy for my daughter to remember and picked her from my family list. I continued cellular service. Then I set up all the usual features and services for an Apple Watch, including Ask to Buy so she couldn’t buy anything from the app store without my permission, Messages, and Emergency SOS.

I also chose to limit my daughter’s contacts on the watch. First, go to Settings > iCloud > Contacts on your phone and make sure it’s toggled on. Then click out, go back to Settings > Screen Time > Family Member > Communication Limits. You need to request your child’s permission to manage their contacts and approve it from the kid’s watch. On their watch, you can add and rename contacts from your contact list (Dad becomes “Grandpa,” Tim becomes “Uncle Timmy,” and so on).

The last step is turning on Schooltime, which is basically a remote-controlled version of an adult Work Focus. It blocks apps and complications, but emergency calls can still come through. The setup tutorial walks you through how to set up Schooltime on your child’s watch, but if you skip it during setup, you can manage it later. On your iPhone, tap All Watches > Your Child’s Watch > Schooltime > Edit Schedule.

I elected to turn Schooltime on when my child is in school and turn it off during afterschool care, but you can also click Add Time if you’d like to turn it on during a morning class, take a break for lunch, and then turn it back on again. Your kid can just turn the digital crown to exit Schooltime, but that’s OK—you can check their Schooltime reports on your iPhone too.

To manage your child’s watch, go to your Watch > All Watches > Family Watches > Your Kid’s Apple Watch. This is how you install updates and manage settings. For more settings that you can turn on or off, check out Apple’s full list here. For example, you can check health details, set up a Medical ID, or even edit their smart replies.

Fun for Everyone

Just as with a grown-up Apple Watch, the first thing you’ll probably want to do is switch the watch face. Hold down the screen and wait for the face to shrink, and swipe to switch. (You probably also want to buy a tiny kid-specific watch band.)

We got my daughter an Apple Watch, so I’d be able to see her on Find My, and she could contact me via phone or the Messages app, which she does with regrettable frequency.