The dream vs. reality
Big Tech companies like OpenAI, Meta, and Google are currently in a hype cycle trying to convince us that their "AI agents" are going to be our ultimate personal assistants. The promise? An AI that can live in your browser, scour the web, book your dinner reservations, and snag the shoes you want without you lifting a finger. Real talk: it’s giving high-level sci-fi energy.
But there’s a massive problem: the actual internet is starting to say "no thanks." In recent weeks, users of Meta’s Muse, xAI’s Grok, OpenAI’s Dots, and the AI agent from Instinct have been running into hard blocks when trying to interact with popular sites.
Who’s blocking who?
It’s not just one company, either. Yelp, Delta Air Lines, JPMorgan Chase, and Charles Schwab have all confirmed they are actively blocking these agents, citing security concerns. Even sites like Cars.com and Amtrak seem to be quietly shutting the door.
Amazon, the king of retail, has been the most aggressive about this, even dragging Perplexity into a legal battle over its AI shopping feature. While Perplexity argued that their agent is just a digital extension of the user, Amazon isn’t vibing with it because—let's be real—bots don't look at ads. Without human eyes on those sponsored products, the revenue model breaks.
The "bot" struggle
Even when sites aren't explicitly banning specific AI models, your agent is likely getting caught in the web’s existing "anti-bot" security systems. These are the same safeguards meant to stop scammers and hackers, and honestly, they can't always tell the difference between a malicious bot and your shiny new AI assistant.
Chris Canal, CEO of EquiStamp, points out that this is a major blow to the perceived value of these agents. If an AI claims it can do everything a human can do online, but it’s constantly getting blocked, then its usefulness is basically zero.
Why it matters
Meta is trying to pivot by making official partnerships with companies like Walmart and Gap, but this whole dynamic is setting us up for a much more annoying web experience. We’re likely headed toward a future where you’ll have to constantly prove you’re human with CAPTCHAs and security checks just to do basic stuff like grocery shopping. Lowkey, the "convenient" future of AI might end up making the internet way harder to navigate.
