Why Websites Are Blocking AI Agents

Major websites are actively blocking autonomous artificial intelligence (AI) agents from completing web tasks on behalf of human users. In recent weeks, users of Meta Muse, xAI Grok, OpenAI Dots, and Instinct have encountered automated roadblocks preventing their software from doing tasks on top websites.

Reported reasons include data-licensing revenue (Yelp), control of bookings (Delta) and account security (JPMorgan Chase, Charles Schwab); Perplexity argues Amazon also wants to protect its ad impressions. In addition, many agent transactions trip automated anti-bot security systems that were originally deployed to prevent malicious scraping, scams, and hacks.

For organizations planning autonomous task execution, navigating websites blocking AI agents requires understanding vendor policies, establishing authorized partner connections, and maintaining human fallbacks when automated browser sessions fail.

Why Websites Are Blocking AI Agents

Websites block AI agents to protect ad revenue (Perplexity's argument about Amazon) and data-licensing revenue, to keep control of bookings and checkout, and over account security, and many agents also trip anti-bot systems built for scrapers; the agents that still work are the ones with a paid or partner arrangement.

According to reporting by Emma Roth in The Verge, users attempting to delegate routine digital tasks to autonomous agents have increasingly encountered hard access blocks.

The core motivations driving web operators to restrict autonomous agents include the following operational factors:

• Advertising protection: Perplexity argues Amazon blocks agents because fewer human eyes see its ads.
• Data licensing enforcement: Platforms like Yelp only allow AI agents that pay for commercial data licensing programs.
• Checkout and booking control: Service providers like Delta Air Lines prevent AI agents from booking tickets on behalf of users.
• Account security defense: Financial institutions like JPMorgan Chase and Charles Schwab block AI agents from accessing accounts, citing security concerns.
• Anti-bot infrastructure: Security systems originally deployed to stop scrapers, scams, and hacks are automatically triggered by AI agents.

Documented Website Blocking Policies

Corporate responses to autonomous agents range from explicit legal bans and technical firewalls to private licensing requirements and commercial distribution partnerships.

Several major brands have confirmed that they block AI agents. Some users say Amtrak and Cars.com, which previously worked with their agent, no longer do. Neither company has confirmed a block.

The table below lists the confirmed policies reported by The Verge and TechCrunch.

The Amazon and Perplexity Dispute

The legal battle between Amazon and Perplexity highlights how e-commerce marketplaces defend retail purchasing paths against autonomous shopping agents.

Amazon banned Perplexity's agentic shopping feature last year and sued Perplexity. Amazon stated that it repeatedly requested Perplexity to halt its shopping agent and insisted that third parties buying goods on behalf of consumers should operate openly and respect service provider decisions on whether to participate. Perplexity argued that Amazon restricts agents because fewer human eyes view its retail advertisements, positioning its shopping agent as a direct digital extension of the consumer.

A judge initially issued an order requiring Perplexity to stop its agents from shopping on Amazon. In August 2026, an appeals court overturned that order, after which Amazon filed an amended complaint. A detailed analysis of the litigation and platform arguments appears in the review of the Amazon v. Perplexity lawsuit.

Amazon said Meta Muse's continued access violates its site policies, and user reports indicate that OpenAI Dots, xAI Grok Bot, and Instinct AI are unable to use Amazon. In response to these retail barriers, Elon Musk publicly remarked that he might consider starting a competitor to Amazon.

Anti-Bot Systems and Automated Security

Many autonomous agents fail on web platforms because their automated navigation routines trip cybersecurity defenses that were never specifically designed to target AI.

Chris Canal, Chief Executive Officer (CEO) of AI safety evaluator EquiStamp, explained to The Verge that anti-bot protections deployed to stop scams and hacks are triggered by autonomous AI agents.

Infrastructure providers have also formalized controls to help site administrators manage synthetic web activity. In 2025, Cloudflare introduced a one-click blocking tool for AI crawlers, and later expanded its mitigation features to block automated bots that scrape online content for both search indexing and AI training.

• Verification hoops: Chris Canal expects more "prove you're human" hoops, even on simple grocery sites.
• Anti-bot triggers: Systems built to stop scams and hacks are triggered by autonomous AI agent traffic.
• Traffic management: Edge networks like Cloudflare block crawlers to manage AI bot traffic straining the web.

Managing Agent Workflows in Your Business

Organizations relying on autonomous agents should design workflows around platform restrictions.

Chris Canal noted that widespread website blocking reduces the perceived utility of autonomous software because there are fewer practical tasks agents can execute. At Layer3Labs, we build custom agents and enterprise integrations for SMB workflows. When selecting tools like Meta Muse or Perplexity Computer, companies must prioritize supported partner ecosystems.

Meta has addressed commercial blocking by establishing authorized partnerships with major retailers including Walmart, Gap, Sephora, and Best Buy, allowing Muse to execute shopping requests through approved channels. Similarly, Shopify has opened its merchant storefront network to automated AI agents. However, user reports on Reddit note unconfirmed checkout problems on Walmart, while reports suggest Google is blocking some accounts that use Muse for "suspicious activity".

To maintain reliable operations despite access restrictions, businesses should implement four concrete operational practices:

• Select authorized merchant channels: Direct automated purchasing tasks toward platforms with explicit agent agreements, such as Shopify merchants or Meta retail partners.
• Verify data licensing options: If an operational workflow depends on directory information from providers like Yelp, budget for official data licensing programs.
• Implement human-in-the-loop checkpoints: Configure autonomous workflows to halt before checkout, payment entry, or account credential screens so a human operator can complete the final verification step.
• Audit internal agent permissions: Review our guide on shadow AI agents risk mitigation to ensure staff do not assign sensitive corporate banking credentials to autonomous browser tools.

Website Operator Decisions on Blocking Agents

Website operators weigh account security, data-licensing revenue and, per Perplexity's argument about Amazon, ad impressions against sales from consumers who shop using automated assistants.

Meta framed this platform tension directly, stating that turning away a personal agent means turning away the human customer who dispatched it. JPMorgan Chase and Charles Schwab cite security concerns for blocking agents from customer accounts.

Yelp is building its own agent while restricting outside agents. For example, TechCrunch reported that Yelp launched Check for Me, an internal voice agent that places phone calls to restaurants on behalf of users to confirm table wait times before texting the results. Yelp also incorporated direct OpenTable restaurant booking capabilities inside its own consumer AI assistant while maintaining strict commercial licensing rules for outside agents.

Operators decide per task whether to turn agents away or admit them under a partner arrangement.

• Commercial revenue trade-off: Preserving direct human ad impressions versus capturing sales volume initiated through autonomous buyers.
• Infrastructure protection: Using Cloudflare rules to block crawlers that scrape for AI training and search indexing.
• Proprietary agent development: Launching internal first-party services like Yelp Check for Me while charging external AI vendors for raw data access.

Auditing Workflow Architecture and Agent Selection

Deciding whether to deploy browser-based agents or traditional API integrations determines how vulnerable a business process remains to third-party access bans.

Teams evaluating workflow automation can consult our breakdown of AI workflow automation vs AI agents to compare both approaches.

When evaluating broader commercial initiatives, enterprise leaders should consult our guides on autonomous AI agents for business and the best AI agents for ecommerce to benchmark supported integrations before deploying automation across external websites.