AI Browsers: Time Savers or Security Dangers?
Most people have grown accustomed to a certain concept of what an internet browser is: it’s a tool for accessing the internet, and that’s it. It isn’t all that smart or technical; it’s more like a window to the stuff I need to see and use.
This definition was never quite right, though we can see how people got there. Most browsers offer a pretty similar experience, so it’s just not something most people give a lot of thought to.
Enter AI: as with nearly every other corner of business technology, AI is showing up inside web browsers. It’s even spawning a new breed of browser built around generative AI from the ground up.
So far, these AI browsers remain more of a niche option. But they’re growing quickly in popularity, so the chances are getting higher by the day that someone wants to use them within your business (or already is).
Here’s what you need to know about AI browsers — what they are, what they’re capable of, and how they could put your business at risk.
What Are AI Browsers?
AI browsers are what they sound like: they’re web browsers with built-in AI capabilities. These range from existing products with AI add-ons, like Microsoft Edge with Copilot, all the way to AI-native tools, like OpenAI’s Atlas and Perplexity’s Comet.
Whatever the level of integration, AI browsers add an AI assistant or chatbot-style experience to your web browsing. Instead of opening up ChatGPT or Perplexity in a new tab, you can access whatever AI model or product is on offer directly from the browser.
What Can AI Browsers Do?
The definition above might feel a little underwhelming. If that were all AI browsers could do, they wouldn’t be all that much better than just opening a separate window and navigating to your GenAI tool of choice.
But that’s not all they can do. Depending on the AI browser you’re evaluating, functions could include:
- Summarizing a long article
- Extracting key points from a PDF or presentation
- Autonomously performing some browsing tasks, like booking a hotel room
- Organizing and prioritizing your open tabs
- Doing light internet research (such as finding the best deal on a product)
Because these functions live in the browser, they happen naturally and intuitively, and you can see the trail of thought or action as it happens.
For some people and certain workflows, AI browsers really do make navigating the internet an easier and more productive process.
What Are the Dangers of Using AI Browsers at Work?
Like any new technology, AI browsers introduce new risks, too. But because AI browsers can autonomously act on your behalf, these risks can feel (and be) a little more significant.
For example, you might already use Copilot or ChatGPT to help you draft an email. Maybe it sharpens up your first draft, or maybe it writes that first draft that you then sharpen up yourself. It’s a little different to hand over the keys to the email inbox, so to speak: AI browsers can navigate to a web page (like Airbnb), write a message on your behalf and send it, all autonomously. Great if that message says what you want and sounds like you. Not so great if that message ends up promising a client stuff you can’t give them.
Other concerns include safety and security: AI agents can get tricked into taking actions you didn’t intend. And because they have access to more of your information, those actions can potentially do more damage.
There’s also the question of data privacy. Any business information you access in an AI browser is something that browser can “see” — creating some level of privacy risk.
Our take: AI browsers are an intriguing development, but unless your team has a specific use case where an AI browser will deliver value, we recommend waiting (or at a minimum sticking to business-oriented versions, like Edge with Copilot).
That’s it for this week. Reach out anytime if you have questions or need assistance.
Comments
Post a Comment