主流网站陆续封锁AI代理 科技巨头落地受阻

AI agents are being broadly blocked by mainstream websites, and regular users need to watch out for their experience and account security.

1. Core phenomenon: AI agents from OpenAI, Meta, Google, xAI and others can perform web searches, restaurant bookings, email cleanup, product purchases and more on behalf of users, but platforms including Yelp, Delta Air Lines, JPMorgan Chase, and Amazon have set up barriers against AI agents. Cars.com, Amtrak, and other previously accessible sites have also become unreachable in succession.

2. Impact on regular users: More websites will likely introduce CAPTCHA or human verification in the future, and daily scenarios like fresh-grocery shopping may require multiple verification rounds. Using AI agents to perform actions can easily trigger anti-bot systems, causing accounts to be mistakenly banned — for instance, Google once banned some accounts that used Muse.

3. Channels still usable: Meta has partnered with Walmart, Gap, Sephora, Best Buy and other retailers to allow Muse to shop on those brand sites. Shopify has also opened its full store network to AI agents. However, even when users shop on partner platforms, they may still encounter system interception at actual checkout.

This article illustrates the channel game between brands and AI agents, as well as the latest partnership models adopted by leading brands.

1. Channel rules: Yelp explicitly grants access only to AI agents that pay for its data licensing program. Amazon blocked Perplexity's agentic shopping feature and filed a lawsuit, primarily to protect ad exposure for human users. Brands must decide for themselves whether to accept AI agents, which directly affects their sales channels and how they reach users.

2. Partnership cases: Meta has reached partnerships with Walmart, Gap, Sephora, Best Buy and other retailers to allow Muse to complete purchases on their brand sites. Shopify has opened its entire store network to AI agents, enabling brands to reach users behind AI agents through these platforms.

3. Product development and user behavior observation: AI agents are changing the entry point for online shopping. In the future, brands may need to optimize product information, checkout flows, and bot-detection rules for AI agents, while avoiding anti-crawler systems that mistakenly harm real users.

This article offers sellers policy interpretation, risk warnings, and channel opportunities, especially regarding Amazon's lawsuit and Shopify's opening move.

1. Platform policy interpretation: Amazon started blocking AI agents as early as last year, banning Perplexity's agentic shopping feature and filing a lawsuit that requires third-party purchasing services to operate transparently and respect service providers' decisions on whether to integrate. An appeals court has since overturned the preliminary injunction, and Amazon has filed an amended complaint. Sellers need to keep watching how the rules evolve.

2. Opportunities: Meta has partnered with Walmart, Gap, Sephora, and Best Buy, and Shopify has opened its entire store network to AI agents. Sellers within Shopify's ecosystem or on partner brand sites may gain new order channels from AI agents.

3. Risk mitigation: AI agent visits may be blocked by anti-bot systems, causing agent-assisted orders to fail. In addition, third-party AI purchasing may violate platform rules, and sellers should also be mindful of distorted ad-performance metrics caused by AI agent traffic.

This article, which covers the blocking and opening-up of AI agents, offers factories insights into channel digitalization and product adaptation.

1. Channel opportunities: Shopify has opened its full store network to AI agents. Factories can use such e-commerce channels to reach large-scale procurement and end-consumer demand behind AI agents, increasing their sales outlets.

2. Product and design requirements: When AI agents make purchases on behalf of users, factories must ensure product information is structured, specifications are clear, and inventory status is accurate, so that AI agents can understand and recommend their products. This affects how data is organized when designing new products.

3. Digitalization insights: The disputes between platforms and AI agents show that e-commerce infrastructure needs clear rules. When connecting with platforms, factories should pay attention to official authorization, data security, and bot-detection requirements, to avoid sales channels being blocked by third-party agent behavior.

This article reflects the industry pain points in AI agent implementation, presenting service providers with technology trends and solution space.

1. Industry trend: AI agents are a hot direction for AI deployment right now, but mainstream websites continue to block them, and the range of usable scenarios keeps narrowing, directly affecting how the market judges the value of products from OpenAI, Anthropic and other AI vendors.

2. Client pain points: Platform bans and site-wide anti-bot systems form double obstacles. For example, Google banned accounts using Muse due to suspicious activity, and even at Walmart, where Meta has a partnership, users can still be intercepted by the system at checkout. Clients urgently need solutions for compliant access, preventing false bans, and account security.

3. Solution references: Yelp uses a paid data-licensing model, Meta chooses collaboration with enterprises to resolve underlying concerns, and Shopify opens its full network to AI agents. Cloudflare has launched a one-click block feature for AI crawlers. Service providers can help brands and AI vendors build authorization management, human-verification, and risk-control solutions.

This article provides platform operators with the latest blocking practices, partnership models, and risk-avoidance references.

1. Latest practices: Yelp grants AI agents access only through a paid data-licensing program. Delta blocks agent-assisted ticket booking. JPMorgan Chase and others prohibit AI agents from accessing accounts for security reasons. Amazon has sued Perplexity and set public operating rules. Cloudflare launched a one-click block for AI crawlers to ease infrastructure pressure and prevent ordinary users' resources from being crowded out.

2. Attracting business and room for cooperation: Meta has partnered with Walmart, Gap, Sephora, and Best Buy, and Shopify has opened its full store network to AI agents. This shows that platforms can attract new user traffic by establishing official collaboration with AI vendors, provided the rules are safeguarded.

3. Operations management and trend avoidance: Platforms need to balance human-user experience with AI traffic, and set up human-verification interfaces to avoid harming ordinary users. Also note that in the Perplexity case, a preliminary injunction was overturned by an appeals court, so platforms should make rules as transparent as possible and leave room for legal adjustments.

This article presents new industry dynamics, legal disputes, and business model innovation around AI agents and website blocking, providing valuable material for research.

1. Industry dynamics: Mainstream websites are blocking AI agents on a large scale. Yelp, Amazon, Delta Air Lines, JPMorgan Chase and others have adopted different strategies, narrowing the scenarios AI agents can cover. This may lead the market to lower its assessment of AI vendors' value and create a new pattern of industry contestation.

2. Legal issues: In the Amazon-Perplexity dispute, a preliminary injunction was overturned by an appeals court. The core controversy is the boundary between platform rules and users' right to hire agents or delegate actions to others. Data licensing, transparent operations, and platform autonomy have become legal focal points, and more research is needed on the legal status of user agency in the platform economy.

3. Business models: Yelp's paid data licensing, Meta's collaboration with retailers, Shopify's opening of its full store network, and Cloudflare's anti-crawler tools constitute new business models and infrastructure contestation in the AI agent era. Researchers should consider their implications for the open internet ecosystem, advertising models, and data governance policy.