The AI browser wars begin
What Perplexity's Comet browser means for AI—and publishers. Plus: Fortune and Axios hire AI writers, YouTube targets AI slop, and more
The next battleground in AI will be the browser, and it could be where the war goes nuclear for the media. Once the omnibox in web browsers—where most searches are done—gives you AI answers by default, it will dramatically reduce the number of people going to websites. Baby steps, though, and one of the first is Perplexity debuting Comet, the first AI browser with a real chance of challenging Chrome.
More on that in a minute, but I'd first like to re-up my excitement about the partnership I announced on Tuesday with my friends at The Upgrade to offer AI training and transformation programs to public relations and communications teams. I knew there would be interest, but I didn't know it would be my most-liked post ever.
To recap, in addition to offering amazing AI training courses for both journalists and PR professionals, our two companies are partnering to work alongside PR firms and corporate comms teams to build out the policies, programs, and workflows to ensure you're getting real results from your investments in AI and staying ahead of your competitors. You can learn more here, and feel free to reach out if you think we can help.
One more plug for my upcoming AI Quick Start class that's happening next week, and then we'll chart where Perplexity's Comet is going.
A MESSAGE FROM THE MEDIA COPILOT
AI isn’t a fad. For most journalists and PR professionals now, it’s simply the reality. Yet many reporters and executives are still stuck dabbling: using ChatGPT for story ideas or a few headlines and calling it a day.
It’s time to level up.
Use the summer to get an unfair advantage: a one-hour power session on how to actually use AI like a pro. Research smarter. Pitch sharper. Write faster—without losing your voice to the algorithm.
Here’s what you’ll walk away with:
The real state of AI in media—no hype, just how it’s changing the game.
A simple way to approach prompting to get what you actually want.
Pro-level use cases: Find better sources. Focus your stories and campaigns. Protect your data.
📅 Live: July 17, 1 p.m. ET
💸 $49: The cheapest upskill you’ll make all year.
🎁 Extras: Slides, prompt templates, and a quickstart AI guide to keep you sharp.
👉 Don’t just survive AI. Use it to do better work.
Save your spot now—before someone else does your job faster.
Perplexity puts an agent in your web browser
One of the best strategic moves Google ever made was launching Chrome. The search engine had already organized the world's information, then it sought to control the browser people used to find it. By targeting pain points in the browser experience—mainly the proliferation of plug-ins that slowed performance—Chrome captured the majority of browser market share (68%, per Statcounter), a title it still holds today.
Now Chrome is the target of disruption. Perplexity, the AI search upstart challenging Google, just launched Comet, an AI-powered browser. And like Chrome before it, it aims to target pain points by adding agent-like behavior to many of the tasks you do every day.
For instance, when I looked up that statistic in my introduction, I opened a new tab, typed "browser market share" in the omnibox, scanned the search results, found a source that looked good, clicked and then found the stat on the page. Or, if I were going to use AI, instead of typing my search I would have opened Gemini or ChatGPT, asked what Chrome's market share was, then gotten the answer directly from chat.
Riding the Comet
With Comet, I could just start typing my query in a new tab, eliminating the step to go anywhere. Early adopters are already doing this via extensions, but Comet does even more with its Comet Assistant: By having access to your search, your activity history, and all the apps you're logged into, Perplexity's browser can know the context of what you're doing. That means you can ask it in natural language to, say, find emails from your family, or any receipts from yesterday. It can even send calendar invites.
The whole idea is to turn the browser into a kind of agent, at least partially, similar to AI extensions like Monica. That's a transformative idea, though it also requires a new level of trust that Perplexity will keep all that context private. As I viewed the demos on Perplexity's Comet page, I was reminded of the headwinds that Microsoft encountered with its AI-powered Recall feature on its Surface PCs back in 2024, which the company had to postpone because it took screenshots every few seconds and creeped people out too much. The Comet Assistant would arguably have an even more intimate picture of your life.
Perplexity is limiting access to subscribers of Perplexity Max, a new $200-a-month subscription tier made to match OpenAI's ChatGPT Pro. Perplexity Pro users at the $20-a-month level can join the waitlist.
Comet isn't the first AI-powered web browser; in June The Browser Company, which created Arc, launched Dia, a browser that turns the omnibox into a chatbot. And it won't be the last, either; OpenAI is reportedly just weeks away from launching its own browser to take on Google Chrome, according to Reuters.
It might not be first to market, but Perplexity is first to have a decent shot at challenging Google. TechCrunch reports it's now fielding 780 million queries a month, and the company is valued at $14 billion. OpenAI has more scale, but Perplexity has been doing search longer, and the fact that it doesn't build its own foundation models may actually be a strength in case of the browser, which people like to customize.
Target: Chrome
Whether it's Perplexity, OpenAI, or Google itself, the Chrome browser and others like it are ripe for disruption. The myriad extensions are starting to feel like the old bloat of plug-ins, and Google has so far not done much to integrate Gemini into the experience. That may be because it's been too focused on building out Gemini, wary of antitrust fallout (Chrome may still be spun off), or simply reluctant to risk its dominant market share.
Regardless, there's an opportunity for someone to bring AI to the browser and knock Chrome off its perch. Making AI answers the default search experience will certainly accelerate all the trends we're seeing with regard to content discovery: dramatically fewer clicks to sites, better engagement from the people who do, and hallucinations creeping even further into the mix.
The evolution of AI so far suggests people will trade some privacy and tolerate a few inaccuracies for the sake of a better experience—one where an LLM is there to help you every step of the way. The AI browser wars have begun. Game on.
A MESSAGE FROM THE MEDIA COPILOT AND THE UPGRADE
AI isn’t here to take your job—it’s here to change how you do it. The smartest media professionals are already using AI to work faster, uncover better stories, pitch more effectively, and stay ahead in an industry that never sits still. But learning how to get real value from AI tools takes more than random experiments and viral prompts.
That’s why we created two focused, 6-week courses with The Upgrade: one for journalists, and one for PR and communications pros. Each program is practical, hands-on, and designed to fit your real work—helping you sharpen your research, write smarter, plan campaigns, and create content with confidence.
For journalists, use code AIJO25OFF-SEP for 25% off until July 31
For PR and comms pros, use code AIPR25OFF-FL1 for 25% off until July 31
The Chatbox
All the AI news that matters to media*
When your co-writer is a machine
Called it. Back in December I predicted that AI-generated content would become more common at major publications, and now Semafor reports that both Fortune and Axios now allow some AI-created copy in stories. In the case of Fortune, they've brought back former editor Nick Lichtenberg to help build Fortune Intelligence, which pairs human editors with AI-generated drafts. For Axios, guidelines that previously prohibited AI-generated stories are being revised, according to media reporter Max Tani. After several infamous forays into AI content, last year ESPN showed it could be done—with the right editorial process. As AI writing gets better and publications continue to get squeezed by the reality of shrinking search referrals, it's going to be hard for any site to resist conscripting AI into a writing role, however reluctantly.
Publishers call out Google's approach to AI crawling
The Independent Publishers Alliance filed an antitrust complaint against Google in the EU, Reuters reports. It shines a light on the core problem in how Google uses AI: publishers can't opt out of having their content used for AI summaries without losing their ability to appear in Google search results entirely. While Google predictably argues that AI Overviews create "new opportunities" for content discovery, the complaint's request for interim measures suggests publishers believe the damage is happening faster than regulators can respond. The case represents the first coordinated pushback against Google's "you wouldn't dare" approach to AI—one that treats publishers as hostages to the existing search ecosystem to buy time to build an AI future that may end up replacing them. (AI-assisted)
Want to be a better reporter? Start an AI diary.
Looking for a better way to teach AI literacy to journalism students? One professor's experiment with "AI diaries"—where students logged every interaction with generative AI tools—produced striking results that go beyond basic transparency. Students who kept these diaries not only showed more critical thinking about AI's limitations but also produced noticeably better journalism. The diary format forced students into what psychologist Daniel Kahneman calls "System 2" thinking: slow, deliberate reflection instead of quick, instinctive responses. Students spent more time developing story ideas, found better sources, and applied course literature throughout their reporting process rather than just quoting it at the end. Perhaps most importantly, both AI enthusiasts and skeptics developed more nuanced views by semester's end, with reduced trust in AI outputs—exactly the kind of critical perspective newsrooms need as they integrate these tools into daily workflows. (AI-assisted)
The BBC carefully and quietly rolls out AI tools
You might be surprised by the BBC's pilot of two AI tools—bullet-point summaries and an automated style assistant—since the news outlet has been critical of AI accuracy in the past. As reported by Forbes, the broadcaster requires human review of every AI output and full disclosure to audiences when AI is used. The Style Assist tool, which reformats partner content into BBC house style, tackles the unglamorous but time-consuming work that newsrooms desperately need to streamline. If the BBC can prove these tools actually save time without compromising accuracy, it would be a big green light for other major outlets to follow suit. (AI-assisted)
YouTube gets out the shovel for AI slop
No one likes slop—not even YouTube. In what it's calling a minor update to its YouTube Partner Program, it's cracking down on creators' ability to monetize "repetitive" and "inauthentic" videos—the kind that overlay AI voices on top of photos or clips, which are often also AI-generated. It's always been junk and not monetizable, according to TechCrunch, but now it's orders of magnitude easier to create. It's a little ironic that YouTube is weeding out AI videos right when it's about to bring its much-lauded Veo 3 video model to the platform, but making distinctions between good, desired AI and bad, garbage AI is something we're all going to have to do going forward. Maybe then AI might even develop its own taste, just as Ben Affleck predicted.
*Some items are AI-assisted. For more on what this means, see this note.