OpenAI, Meta, and Google want AI agents to do everything for you. But popular services on the web aren\u2019t letting that happen.<\/h5>\n\n\n\n![\"STK414_AI_CVIRGINIA_C\"\/](\"https:\/\/platform.theverge.com\/wp-content\/uploads\/sites\/2\/2025\/03\/STK414_AI_CVIRGINIA_C.jpg?quality=90&strip=all&crop=0%2C0%2C100%2C100&w=2400\")
Image: Cath Virginia \/ The Verge, Getty Images<\/figcaption><\/figure>\n\n\n\nBy Emma Roth<\/p>\n\n\n\n
Autonomous agents are the hottest thing in AI. OpenAI, Meta, Google, Instinct, and others promise their agents can scour the internet on your behalf, doing anything from booking a dinner reservation and cleaning out your inbox to buying new shoes. But what happens when the rest of the internet won\u2019t play nice?<\/p>\n\n\n\n
In recent weeks, users of services like Meta\u2019s Muse, xAI\u2019s Grok, OpenAI\u2019s Dots, and Instinct have reported blocks preventing their AI agents from performing tasks across top websites. Yelp told TechCrunch<\/em><\/a> that it only allows AI agents that pay for its data licensing program, while Delta Air Lines is preventing them from booking tickets on behalf of users. JPMorgan Chase<\/a> and Charles Schwab<\/a> have also confirmed that they\u2019re blocking AI agents from accessing accounts due to security concerns. Some users say sites that previously worked with their AI agent are no longer accessible, including Cars.com<\/a> and Amtrak<\/a>, suggesting other companies could be quietly blocking them.<\/p>\n\n\n\nAmazon was already fighting off AI agents last year when it banned Perplexity\u2019s agentic shopping<\/a> feature and sued the AI startup<\/a>, raising legal and practical questions about the future of agents on the web. At the time, Amazon claimed it \u201crepeatedly requested\u201d that Perplexity stop allowing its AI agents to shop on its website, adding that third-party companies that purchase items on behalf of customers \u201cshould operate openly and respect service provider decisions whether or not to participate.\u201d<\/p>\n\n\n\nPerplexity responded<\/a> by pointing out that Amazon doesn\u2019t want AI agents roaming its platform because it won\u2019t get as many human eyeballs in front of its ads \u2014 which is<\/em> an actual issue<\/a> for the ecommerce giant. The AI startup likened its AI agent to an extension of the user, arguing that Amazon \u201cdoes not believe in your right to hire labor, to have an assistant or an employee acting on your behalf.\u201d While a judge later ordered Perplexity<\/a> to stop its AI agents from shopping on its website, an appeals court overturned this ruling in August<\/a>. Amazon has since filed an amended complaint<\/a>.<\/p>\n\n\n\nEven if websites aren\u2019t banning AI agents outright, agents may still be clashing with anti-bot systems<\/p>\n\n\n\n
This ongoing legal battle didn\u2019t stop OpenAI, Meta, and Google from rolling out agents of their own. But they\u2019re dealing with the same problems as Perplexity. Muse, Meta\u2019s deceptively cute AI assistant, most recently faced the wrath of Amazon<\/a>, which said its \u201ccontinued access\u201d violates its site policies. There are reports of xAI\u2019s Grok<\/a>, Instinct\u2019s AI agent,<\/a> and OpenAI\u2019s Dots<\/a> being unable to use Amazon as well. xAI owner Elon Musk responded<\/a> by saying that maybe he \u201cshould start a competitor to Amazon.\u201d The more places bots struggle to access, the greater the questions about how useful they are.<\/p>\n\n\n\n\u201cDoes this impact the perceived value of the Anthropics and the OpenAIs? I would say absolutely,\u201d Chris Canal, the CEO of the AI safety evaluator EquiStamp, tells The Verge<\/em>. \u201cOne of the claims is that they\u2019re going to permeate every aspect of our lives and be able to do anything that we could do online. If there are companies that are actively working to make that not happen, there are just fewer useful things the agents can do.\u201d<\/p>\n\n\n\nMeta responded to instances of websites blocking Muse in a post earlier this month<\/a>. \u201cTurning away a personal agent means turning away the customer behind it,\u201d the company says. \u201cWe\u2019d rather work with businesses to address the underlying concerns than see them lose that customer.\u201d Instead of trying to fight Amazon, the company has partnered with retailers<\/a> like Walmart, Gap, Sephora, Best Buy, and others, allowing Muse to shop on their sites. At the same time, Shopify has also opened its network of storefronts<\/a> to AI agents.<\/p>\n\n\n\nThe Verge<\/em> reached out to OpenAI and Google with requests for comments on the roadblocks their AI agents are facing but didn\u2019t immediately hear back. Instinct declined to comment.<\/p>\n\n\n\nEven if websites aren\u2019t banning AI agents outright, agents may still be clashing with anti-bot systems, which have become ubiquitous across the web. Reports on Reddit<\/a> suggest that Google is blocking some accounts that use Muse due to \u201csuspicious activity,\u201d while others have reported issues<\/a> checking out on Walmart despite its Meta partnership. \u201cThere are all kinds of things that are set up not even to block agents, but to block in general bots or programs that are trying to scam or hack,\u201d Canal adds. \u201cThese AI agents trigger those same safeguards.\u201d<\/p>\n\n\n\nThe web architecture provider Cloudflare has also played a role in helping sites clamp down on unwanted AI visitors. After introducing a one-click block for AI crawlers<\/a> in 2025, Cloudflare began blocking crawlers<\/a> that simultaneously scrape websites for both search indexing and AI training. The move is part of Cloudflare\u2019s efforts to save the web from an influx of AI bots, which is putting a strain on internet infrastructure<\/a> and wreaking havoc on email inboxes, and which could eventually make it even harder<\/a> than it already is to compete against bots when purchasing concert tickets, booking reservations, or trying to get ahold of certain products.<\/p>\n\n\n\nWhile these blocks may help keep some AI agents and bots out of a website, Canal says they will ultimately make the web more \u201cannoying\u201d to use as more sites roll out interfaces to confirm you\u2019re not a robot. \u201cWe could see them on a simple grocery shopping website soon, where you\u2019re having to jump through a bunch of hoops to prove that you\u2019re a human using the computer or your phone.\u201d<\/p>\n","protected":false},"excerpt":{"rendered":"
I heard someone on a podcast talking about Muse specifically, that when blocked it asks for permission to use your browser to look like you browsing, so they’re getting you to give up more access to work around the issue. I don’t mess with this stuff myself, as it’s mostly spyware to build a very […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6,7],"tags":[],"class_list":["post-19134","post","type-post","status-publish","format-standard","hentry","category-tech","category-world"],"featured_image_src":null,"author_info":{"display_name":"Jason","author_link":"https:\/\/jasonsblog.ddns.net\/index.php\/author\/jturning\/"},"_links":{"self":[{"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/posts\/19134","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/comments?post=19134"}],"version-history":[{"count":1,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/posts\/19134\/revisions"}],"predecessor-version":[{"id":19135,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/posts\/19134\/revisions\/19135"}],"wp:attachment":[{"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/media?parent=19134"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/categories?post=19134"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/tags?post=19134"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
By Emma Roth<\/p>\n\n\n\n
Autonomous agents are the hottest thing in AI. OpenAI, Meta, Google, Instinct, and others promise their agents can scour the internet on your behalf, doing anything from booking a dinner reservation and cleaning out your inbox to buying new shoes. But what happens when the rest of the internet won\u2019t play nice?<\/p>\n\n\n\n
In recent weeks, users of services like Meta\u2019s Muse, xAI\u2019s Grok, OpenAI\u2019s Dots, and Instinct have reported blocks preventing their AI agents from performing tasks across top websites. Yelp told TechCrunch<\/em><\/a> that it only allows AI agents that pay for its data licensing program, while Delta Air Lines is preventing them from booking tickets on behalf of users. JPMorgan Chase<\/a> and Charles Schwab<\/a> have also confirmed that they\u2019re blocking AI agents from accessing accounts due to security concerns. Some users say sites that previously worked with their AI agent are no longer accessible, including Cars.com<\/a> and Amtrak<\/a>, suggesting other companies could be quietly blocking them.<\/p>\n\n\n\n Amazon was already fighting off AI agents last year when it banned Perplexity\u2019s agentic shopping<\/a> feature and sued the AI startup<\/a>, raising legal and practical questions about the future of agents on the web. At the time, Amazon claimed it \u201crepeatedly requested\u201d that Perplexity stop allowing its AI agents to shop on its website, adding that third-party companies that purchase items on behalf of customers \u201cshould operate openly and respect service provider decisions whether or not to participate.\u201d<\/p>\n\n\n\n Perplexity responded<\/a> by pointing out that Amazon doesn\u2019t want AI agents roaming its platform because it won\u2019t get as many human eyeballs in front of its ads \u2014 which is<\/em> an actual issue<\/a> for the ecommerce giant. The AI startup likened its AI agent to an extension of the user, arguing that Amazon \u201cdoes not believe in your right to hire labor, to have an assistant or an employee acting on your behalf.\u201d While a judge later ordered Perplexity<\/a> to stop its AI agents from shopping on its website, an appeals court overturned this ruling in August<\/a>. Amazon has since filed an amended complaint<\/a>.<\/p>\n\n\n\n Even if websites aren\u2019t banning AI agents outright, agents may still be clashing with anti-bot systems<\/p>\n\n\n\n This ongoing legal battle didn\u2019t stop OpenAI, Meta, and Google from rolling out agents of their own. But they\u2019re dealing with the same problems as Perplexity. Muse, Meta\u2019s deceptively cute AI assistant, most recently faced the wrath of Amazon<\/a>, which said its \u201ccontinued access\u201d violates its site policies. There are reports of xAI\u2019s Grok<\/a>, Instinct\u2019s AI agent,<\/a> and OpenAI\u2019s Dots<\/a> being unable to use Amazon as well. xAI owner Elon Musk responded<\/a> by saying that maybe he \u201cshould start a competitor to Amazon.\u201d The more places bots struggle to access, the greater the questions about how useful they are.<\/p>\n\n\n\n \u201cDoes this impact the perceived value of the Anthropics and the OpenAIs? I would say absolutely,\u201d Chris Canal, the CEO of the AI safety evaluator EquiStamp, tells The Verge<\/em>. \u201cOne of the claims is that they\u2019re going to permeate every aspect of our lives and be able to do anything that we could do online. If there are companies that are actively working to make that not happen, there are just fewer useful things the agents can do.\u201d<\/p>\n\n\n\n Meta responded to instances of websites blocking Muse in a post earlier this month<\/a>. \u201cTurning away a personal agent means turning away the customer behind it,\u201d the company says. \u201cWe\u2019d rather work with businesses to address the underlying concerns than see them lose that customer.\u201d Instead of trying to fight Amazon, the company has partnered with retailers<\/a> like Walmart, Gap, Sephora, Best Buy, and others, allowing Muse to shop on their sites. At the same time, Shopify has also opened its network of storefronts<\/a> to AI agents.<\/p>\n\n\n\n The Verge<\/em> reached out to OpenAI and Google with requests for comments on the roadblocks their AI agents are facing but didn\u2019t immediately hear back. Instinct declined to comment.<\/p>\n\n\n\n Even if websites aren\u2019t banning AI agents outright, agents may still be clashing with anti-bot systems, which have become ubiquitous across the web. Reports on Reddit<\/a> suggest that Google is blocking some accounts that use Muse due to \u201csuspicious activity,\u201d while others have reported issues<\/a> checking out on Walmart despite its Meta partnership. \u201cThere are all kinds of things that are set up not even to block agents, but to block in general bots or programs that are trying to scam or hack,\u201d Canal adds. \u201cThese AI agents trigger those same safeguards.\u201d<\/p>\n\n\n\n The web architecture provider Cloudflare has also played a role in helping sites clamp down on unwanted AI visitors. After introducing a one-click block for AI crawlers<\/a> in 2025, Cloudflare began blocking crawlers<\/a> that simultaneously scrape websites for both search indexing and AI training. The move is part of Cloudflare\u2019s efforts to save the web from an influx of AI bots, which is putting a strain on internet infrastructure<\/a> and wreaking havoc on email inboxes, and which could eventually make it even harder<\/a> than it already is to compete against bots when purchasing concert tickets, booking reservations, or trying to get ahold of certain products.<\/p>\n\n\n\n While these blocks may help keep some AI agents and bots out of a website, Canal says they will ultimately make the web more \u201cannoying\u201d to use as more sites roll out interfaces to confirm you\u2019re not a robot. \u201cWe could see them on a simple grocery shopping website soon, where you\u2019re having to jump through a bunch of hoops to prove that you\u2019re a human using the computer or your phone.\u201d<\/p>\n","protected":false},"excerpt":{"rendered":" I heard someone on a podcast talking about Muse specifically, that when blocked it asks for permission to use your browser to look like you browsing, so they’re getting you to give up more access to work around the issue. I don’t mess with this stuff myself, as it’s mostly spyware to build a very […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6,7],"tags":[],"class_list":["post-19134","post","type-post","status-publish","format-standard","hentry","category-tech","category-world"],"featured_image_src":null,"author_info":{"display_name":"Jason","author_link":"https:\/\/jasonsblog.ddns.net\/index.php\/author\/jturning\/"},"_links":{"self":[{"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/posts\/19134","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/comments?post=19134"}],"version-history":[{"count":1,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/posts\/19134\/revisions"}],"predecessor-version":[{"id":19135,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/posts\/19134\/revisions\/19135"}],"wp:attachment":[{"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/media?parent=19134"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/categories?post=19134"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/jasonsblog.ddns.net\/index.php\/wp-json\/wp\/v2\/tags?post=19134"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}