Why it matters
Parsers discard the whole block, so the page loses every entity, breadcrumb and article signal in it. This is a scoring gate.
How to fix it
Generate JSON-LD with JSON.stringify from an object, never by string concatenation, and escape < (see example).
Example
// components/json-ld.tsx (Server Component). JSON.stringify doesn't escape "<", so a string containing
// </script> would break out of the tag; replace it with the JSON escape \u003c.
import type { Graph, Thing, WithContext } from 'schema-dts'
export function JsonLd({ data }: { data: WithContext<Thing> | Graph }) {
return (
<script
type="application/ld+json"
dangerouslySetInnerHTML={{ __html: JSON.stringify(data).replace(/</g, '\\u003c') }}
/>
)
}References
More ai search rules
void geo reports 41 rules in this category. Generative-engine optimisation: whether AI crawlers that don't run JavaScript see the same content, valid and visible JSON-LD, an explicit AI robots policy, llms.txt, Markdown mirrors and answer-first writing.
js-dependencyMain content only appears after JavaScript runshidden-streamed-contentCrawlers receive main content only inside hidden streaming containersai-bot-blockedAI crawler user agents get 401/403/429/503 or a challenge pageai-content-differsAI crawlers get materially different HTML than browsersjsonld-unsafeJSON-LD contains a literal </script or <!--jsonld-missingPage has no JSON-LDjsonld-not-in-rawJSON-LD is injected by JavaScripthome-entity-missingHome page lacks Organization + a single WebSite node