AI Agents Book Appointments and the Pentagon Gets ChatGPT | NCR

AI agents just crossed a line that matters for anyone running a business: OpenAI’s ChatGPT can now sign into websites and complete tasks like booking appointments on your behalf, while the Pentagon has onboarded 1.7 million of its 3 million personnel onto its GenAI.mil platform in under a year. Meanwhile, the Microsoft executive in charge of Office is publicly telling people to stop sending AI-generated documents to colleagues. Three stories, one direction of travel.

Can ChatGPT actually book appointments for you now?

Yes. OpenAI updated ChatGPT Work, its agent mode for autonomous tasks, so the AI can sign into websites and complete multi-step actions: booking a DMV appointment, canceling a reservation, filling out a job application. According to The Verge’s coverage, OpenAI says it does this without the company ever seeing your login credentials.

That’s a real jump from “type a prompt, get text back.” The AI now handles authenticated web interactions, the kind of tasks that used to require you sitting in front of a screen. If you’ve been building automations with Make or Zapier for form submissions and bookings, the calculus changed; the AI doesn’t need your automation anymore. It just does the task itself.

I covered how AI agents are becoming employees and what ChatGPT Work’s autonomy means for solo builders. The pattern keeps accelerating: tools are moving from “suggests what to do” to “does it without asking.”

One thing to watch. If ChatGPT can sign into websites, what happens when it signs into the wrong ones, or fills out forms incorrectly? The Cursor situation, where OpenAI cut off model access, shows platform control is real. When the AI holds your credentials, that control extends to your accounts.

Why did the Pentagon put ChatGPT and Grok on its own platform?

The Department of Defense added custom versions of ChatGPT and Grok to GenAI.mil, its portal for unclassified military AI use. Per the Pentagon’s release, the portal launched last year with Google’s Gemini and has already onboarded more than 1.7 million of the department’s 3 million personnel.

The military angle is the least interesting part. What matters is adoption speed: an institution famous for bureaucracy got 1.7 million people onto an AI platform in under a year. When that’s the benchmark, “we’re still evaluating AI” stops being a defensible position in a company of ten people.

If you’re a solo builder wondering whether AI tools are ready for business-critical work, the US military has already answered. The question is whether you’re adopting fast enough, not whether to adopt.

Why is Microsoft’s Office chief complaining about AI slop?

Because he received one. Ryan Roslansky, who leads Microsoft’s Office organization and oversees LinkedIn, posted a warning on LinkedIn about AI-generated slop showing up in daily work. A colleague sent him a fully AI-generated document, and he wasn’t happy about it. LinkedIn, which recently added an AI slop reporting button, has seen over a million people use it.

Of the three stories this week, this one matters most if you use AI tools professionally. The problem isn’t that AI can’t write; it’s that people are using it to skip thinking through things themselves. A fully AI-generated document isn’t a productivity boost. It’s a signal that someone isn’t doing their job.

I’ve written about the AI groupthink problem and how solo builders can avoid it. The tool is powerful; the failure mode is laziness. If your AI output looks like everyone else’s, you’ve turned your business into a commodity.

The practical takeaway: use AI for research, first drafts, data analysis, and automation. Don’t use it to replace thinking. The exec who builds AI tools for a living is telling you what your gut already knows — if a document reads like AI wrote it, AI shouldn’t have written it.

What else happened this week?

Google wants to license Disney characters for AI tools. Google is pitching studios on AI for production and visual effects, but unions and creators are pushing back over copyright and job displacement. If it goes through, content production at scale changes.

Google AI Mode will book hotels. Rolling out in the US in the coming weeks, with support for Expedia, Booking.com, Hilton, Marriott, and others. You’ll also be able to track flights in points by linking airline rewards. The AI tool ecosystem keeps expanding into real-world transactions.

Amazon is adding 2 million more Nvidia GPUs to AWS, on top of the 1 million already planned, because demand “exceeded expectations.” Nvidia posted record revenue again. The buildout isn’t slowing, which means the tools you use keep getting faster and cheaper.

OpenAI’s advertising business hit a $1B annualized run rate, roughly 200 days after launch. OpenAI is also testing outcome-based pricing that charges only when AI agents succeed. If that model spreads, it changes how you evaluate AI tool costs.

What should you actually do about all this?

Adopt the agents, skip the slop. ChatGPT signing into websites and the Pentagon deploying AI to 1.7 million people both point the same direction: AI is moving from “tool I use” to “agent that acts on my behalf.” Start with the AI Tool Advisor if you’re picking tools, or build your first automation if you’re just getting started. And when the output reads like AI wrote it, rewrite it yourself.

FAQs

Can ChatGPT sign into websites and book appointments? Yes. OpenAI’s ChatGPT Work agent mode can now sign into websites and complete tasks like booking a DMV appointment, canceling a reservation, or filling out a job application. OpenAI says it handles the logins without the company ever seeing your credentials, which moves agents from suggesting actions to completing authenticated, multi-step tasks on your behalf.

What is GenAI.mil? GenAI.mil is the Department of Defense’s portal for unclassified military use of AI tools. It launched last year with Google’s Gemini and now includes custom versions of ChatGPT and Grok. More than 1.7 million of the department’s 3 million personnel have already onboarded, according to the Pentagon’s release.

What is AI slop and why does it matter? AI slop is low-effort, fully AI-generated content sent as if it were real work. Microsoft’s Office chief Ryan Roslansky warned about it after receiving an AI-generated document from a colleague, and LinkedIn’s slop reporting button has been used by over a million people. The fix: use AI for drafts and research, but do the thinking yourself.

Is AI adoption moving fast enough to justify adopting now? The Pentagon onboarded 1.7 million people to an AI platform in under a year, and OpenAI’s ad business hit a $1B annualized run rate in about 200 days. When institutions that slow by reputation move that quickly, waiting to “evaluate” AI tools is a bigger risk than adopting them.