A disturbing new attack vector is emerging in the wild, and it turns out the culprit might be sitting right on your website's homepage. Security researchers have documented how threat actors are exploiting website search boxes as springboards for SEO poisoning campaigns that target AI Overview summaries—ultimately redirecting users to scammer-controlled phone numbers instead of legitimate customer service lines.

The Royal Caribbean Fiasco

Last August, a man planning a cruise googled Royal Caribbean's customer service number. Google's AI Overview dutifully served up what appeared to be the correct contact information at the top of its results. He called it, handed over his credit card details, and the number belonged to scammers. Similar incidents plagued searches for Southwest Airlines customer service numbers, with AI-generated summaries pointing users directly into fraudsters' hands.

How Search Boxes Become Weaponized

The attack exploits a fundamental assumption: that search queries on legitimate websites generate indexable content. Attackers craft specific long-tail search queries—usually high-intent phrases like "customer service phone number" or "refund request contact"—targeting sites with vulnerable search implementations. When Google's crawlers index these results and AI Overview synthesizes its summary, malicious entries can slip into what users perceive as trusted recommendations.

The Developer Security Gap

From an infrastructure perspective, this exposes a critical blind spot. Most development teams focus on securing input fields against SQL injection or XSS but overlook the search box's role in your site's external attack surface. Your search functionality is essentially a public-facing API—and like any endpoint, it needs rate limiting, query sanitization, and monitoring for abuse patterns.

What You Can Do About It

Several defensive measures can reduce exposure: implement aggressive rate limiting on search endpoints to prevent automated abuse; add CAPTCHA or similar friction for repeated queries from the same session; monitor your search logs for known scammer targeting patterns like phone number requests; and consider implementing robots.txt restrictions on search result pages if they're not meant to be indexed.

Key Takeaways

  • Attackers are weaponizing website search boxes to inject malicious results into AI Overview summaries, redirecting users to scammer hotlines.
  • High-profile targets like Royal Caribbean and Southwest Airlines have seen their customers fall victim to AI-synthesized scam numbers.
  • The root cause is that many sites don't protect search endpoints against automated abuse or prevent indexing of search result pages.
  • Treat your site's search functionality as production infrastructure: implement rate limiting, CAPTCHA friction, monitoring, and robots.txt restrictions.

The Bottom Line

This isn't theoretical—real people are losing real money because AI summaries trust content that slipped through vulnerable search implementations. If you're running a site with a public-facing search box, treat it like production infrastructure with proper hardening, monitoring, and incident response plans.