Within a month of each other this year, two companies that shape how most people interact with the internet announced the same idea.
At Google I/O in May, Google introduced information agents that watch the open web on your behalf and notify you when something relevant changes. At its Worldwide Developers Conference in June, Apple announced Notify Me, a Safari feature that watches a single page you choose and alerts you when it updates.
Google’s head of Search, Liz Reid, said from the keynote stage that “we’re entering the era of search agents,” and that phrase captures a real change in how we’re going to be finding things online faster and personalized.
Search Is Shifting From Answers to Ongoing Awareness
For most of its history, search has been built around a single moment of intent. You have a question, you type it in, you get an answer, and the exchange is over. That works well for one-off questions, but much of what people want to know is constantly changing. People want to know when a product comes back in stock, when a price drops, when tickets go on sale, when an appointment opens, or when a policy quietly changes.
Take a few ordinary cases. A parent reloads a summer-camp registration page for days waiting for sign-ups to open, because the best camp spots fill within minutes. A shopper checks a sold-out sneaker product page every morning, hoping their size returns. A compliance team checks a regulator’s website every week, watching for the guidance change that could catch them off guard.
All of these come down to the same frustration. Whatever you’re waiting for is sitting on a page somewhere, so you check it, refresh it, and check it again, and the update still seems to land the moment you finally look away. So the next phase of search adds a second kind of request to the familiar one. Beyond “find this for me,” people increasingly want “watch this for me and tell me when it matters.”
AI Improves on the Original Alerts Model
The idea of software watching the web for you has been around for years. Google Alerts, which launched in 2003, was an early and widely used version of the idea, where you entered a keyword and Google emailed you when it indexed new matching content.
Reporters, researchers, communications teams, and plenty of people have relied on it for more than two decades. The newest agents are built to run continuously in the background and reason across many sources, but the lineage from those simple keyword alerts is still the clearest way to see what has actually changed.
What is new is the layer of judgment that AI adds on top. Keyword alerts are good at flagging that something appeared, but they can’t tell you if something was removed, updated or whether the change meets the condition you actually cared about.
An AI agent can read a request written in plain language, look across different kinds of sources, filter out the noise, and recognize whether the specific thing you asked about has happened. The move from alerts to agents is a move from detection to relevance, where the system understands not just that a page changed but whether the change is one worth notifying you for.
Discovering the Web and Watching a Page Are Different Jobs
These tools work in one of two ways. Some search broadly across the open web, taking a standing question and checking blogs, news, social posts, and data feeds for anything relevant, which works when you do not know in advance where the answer will appear. Others monitor a single page, watching one specific page you have already chosen and catching changes on it, which works when you know exactly which listing or page matters.
Google’s information agents lean toward the first approach and Apple’s Notify Me toward the second, and increasingly the two work together, with a broad agent handing a single page to a monitor to keep the watch. They are not alone, either. OpenAI has pushed ChatGPT in the same proactive direction, running research in the background and delivering briefings without being asked each time.
For Consumers and Businesses, the Win Is Less Noise and Better Timing
It would be easy to assume this is just about getting more notifications. The real value runs the other way. These agents aren’t meant to ping you more often, they’re meant to keep you from missing the things that matter. Search agents cut how often you have to check pages manually, so always catch important changes as they happen.
The use cases for monitoring are virtually limitless. Individuals watch for restocks, price drops, event tickets going on sale, and appointment slots opening up. Businesses watch for regulatory updates, compliance changes, and competitor moves, where finding out a few days late can carry a real cost.
The pandemic showed how much timing matters, when people used monitoring tools to grab scarce COVID vaccine appointments and online grocery delivery slots that vanished in seconds. The same logic drives business, where teams watch competitor pricing, public tenders, and job postings, and where finding out a few days late can carry a real cost.
This is also why watching everything is the wrong goal. An agent that monitors too much simply recreates the noise it was meant to remove, burying the one update that mattered under dozens that did not. The agents people keep coming back to will be the ones that understand a request well enough to stay quiet most of the time. Judging what deserves an interruption is the harder and more useful skill, and it’s where AI earns its place.
The Web Will Be Built to Be Watched
The signal in the Google and Apple announcements is that search is becoming more persistent, more agentic, and more personal. For most of its life, the mark of a great search experience was how fast it could answer a question. Increasingly the mark will be how intelligently it keeps watch for the things you do not want to miss and surfaces them while you can still act. As agents like these become an ordinary part of how we browse and search, the web will increasingly be built not only to be found, but to be monitored.
