{"id":780,"date":"2026-06-29T03:55:26","date_gmt":"2026-06-29T03:55:26","guid":{"rendered":"https:\/\/maxaeo.ai\/blog\/ai-search-raci\/"},"modified":"2026-06-29T03:55:26","modified_gmt":"2026-06-29T03:55:26","slug":"ai-search-raci","status":"publish","type":"post","link":"https:\/\/maxaeo.ai\/blog\/ai-search-raci\/","title":{"rendered":"AI Search RACI: Who Handles Wrong Answers, Bad Citations, and Lost Recommendations?"},"content":{"rendered":"<p>An <strong>AI search RACI<\/strong> is a responsibility matrix that names who diagnoses, who approves the fix, who updates the source, and who reports the business risk every time ChatGPT, Gemini, Perplexity, or Google AI Overviews gets your brand wrong. Without one, a wrong answer pings five Slack channels and gets fixed by nobody.<\/p>\n<p>Most marketing teams already track AI visibility. Far fewer have decided <strong>who acts when the tracker turns red<\/strong> \u2014 and that gap is where lost recommendations quietly compound, because &quot;everyone&#39;s job&quot; is no one&#39;s job.<\/p>\n<p>This guide gives you one original artifact: a complete AI search RACI scoped to three failure types \u2014 wrong answers, bad citations, and lost recommendations \u2014 and the four jobs each one creates. It maps real AI search incidents to the marketing, PR, content, and technical roles that resolve them, plus a worked incident walkthrough and severity tiers you can copy.<\/p>\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" style=\"max-width:100%;height:auto\" loading=\"lazy\"  src=\"https:\/\/maxaeo.ai\/blog\/wp-content\/uploads\/2026\/06\/1782474437826-5-37831-1.jpg\" alt=\"AI search RACI matrix mapping roles to wrong answers, bad citations, and lost recommendations\"><\/figure>\n<h2>What is an AI search RACI?<\/h2>\n<p><strong>An AI search RACI is a responsibility assignment matrix for AI search incidents.<\/strong> It tags every task \u2014 detecting a wrong answer, approving a corrected fact, repairing a citation, reporting pipeline risk \u2014 with one of four roles: <strong>R<\/strong>esponsible (does the work), <strong>A<\/strong>ccountable (owns the outcome, one person only), <strong>C<\/strong>onsulted (gives input), <strong>I<\/strong>nformed (kept in the loop).<\/p>\n<p>The letters are borrowed from classic project management. What&#39;s new is the <strong>column and row set<\/strong>. Instead of &quot;design mockups&quot; or &quot;QA the build,&quot; the rows are AI-search-specific failures, and the columns are the roles that own brand accuracy inside answer engine optimization and generative engine optimization work.<\/p>\n<p>The point is speed and clarity. When an AI assistant misstates your pricing, the matrix answers the only question that matters in the first hour: <em>who has the pen, and who signs off?<\/em><\/p>\n<h2>Why AI search needs its own RACI<\/h2>\n<p><strong>AI search needs a dedicated RACI because the failure modes don&#39;t map to any existing owner.<\/strong> A wrong AI answer is part content problem, part PR problem, part technical problem, and part data problem \u2014 so it falls between desks unless you pre-assign it.<\/p>\n<p>The accuracy problem is not rare. When the <a href=\"https:\/\/www.cjr.org\/tow_center\/we-compared-eight-ai-search-engines-theyre-all-bad-at-citing-news.php\" target=\"_blank\" rel=\"noopener\">Tow Center for Digital Journalism tested eight AI search tools across 1,600 news queries<\/a>, the tools returned incorrect citations in more than 60% of cases \u2014 with error rates running from 37% (Perplexity) to 94% (Grok-3). When your brand is the subject of that error, the cost isn&#39;t an academic footnote \u2014 it&#39;s a buyer reading the wrong thing at the moment of decision.<\/p>\n<p>Three things break without an AI search RACI:<\/p>\n<ul>\n<li><strong>Detection without action.<\/strong> Your ai visibility tool flags a drop, the alert lands in a shared channel, and three people assume the fourth is handling it.<\/li>\n<li><strong>Fixes without sign-off.<\/strong> A content writer &quot;corrects&quot; a claim that PR had deliberately worded a certain way, and the two versions now contradict each other across owned and earned sources.<\/li>\n<li><strong>Risk without a number.<\/strong> Leadership hears &quot;AI is saying weird stuff about us&quot; instead of &quot;we lost 18% ai share of voice on three buying-intent prompts.&quot;<\/li>\n<\/ul>\n<p>A RACI converts vague anxiety into a routed, owned, measurable workflow. It is the operating layer on top of ai search monitoring: the tracker tells you <em>what<\/em> changed; the RACI decides <em>who moves<\/em>.<\/p>\n<h2>The four jobs every AI visibility incident creates<\/h2>\n<p><strong>Every AI search incident, no matter the type, spawns four distinct jobs.<\/strong> Naming them is what makes the matrix assignable. Confuse them and you get the classic failure: the person who <em>spotted<\/em> the problem is assumed to <em>own<\/em> it.<\/p>\n<ol>\n<li><strong>Diagnose<\/strong> \u2014 confirm the issue is real, reproduce it across assistants, and find the source the model is pulling from. This is detection plus root cause, not a glance at one screenshot.<\/li>\n<li><strong>Approve source changes<\/strong> \u2014 decide what the <em>correct<\/em> canonical statement is. This is a brand-truth decision, not a writing task, and it needs a single accountable owner.<\/li>\n<li><strong>Update content<\/strong> \u2014 change owned pages and repair off-site sources so the model has accurate material to retrieve on the next crawl.<\/li>\n<li><strong>Report business risk<\/strong> \u2014 translate the incident into revenue, reputation, and pipeline exposure for leadership, and track whether the fix held.<\/li>\n<\/ol>\n<p>These four jobs need <strong>different skills and different authority<\/strong>. Diagnosis is analytical, approval is strategic, updating is executional, reporting is financial. One person rarely owns all four well \u2014 which is exactly why a matrix beats a hero. To separate &quot;fix now&quot; from &quot;log for later&quot; before you assign jobs, pair this with a structured way of <a href=\"https:\/\/maxaeo.ai\/blog\/ai-search-issue-triage\">ranking AI search problems by revenue, reputation, and fixability<\/a>.<\/p>\n<h2>The roles in an AI search RACI<\/h2>\n<p><strong>Seven roles cover almost every AI search incident in a B2B marketing org.<\/strong> Collapse them in a small team or split them in a large one \u2014 the functions stay constant.<\/p>\n<ul>\n<li><strong>GEO\/AEO Lead<\/strong> \u2014 owns the AI search program end to end; the default incident commander.<\/li>\n<li><strong>Content Owner<\/strong> \u2014 owns on-site pages, docs, and the words search assistants quote.<\/li>\n<li><strong>Technical SEO \/ Web<\/strong> \u2014 owns crawlability, schema, and whether assistants can even read your site.<\/li>\n<li><strong>PR &amp; Comms<\/strong> \u2014 owns earned sources: press, third-party reviews, and brand mentions in ChatGPT-cited media.<\/li>\n<li><strong>Product \/ Brand Marketing<\/strong> \u2014 owns positioning truth and what counts as a &quot;correct&quot; claim.<\/li>\n<li><strong>Data \/ Analytics<\/strong> \u2014 owns measurement: mention rate, citations, and ai share of voice trends.<\/li>\n<li><strong>Executive Sponsor<\/strong> \u2014 owns budget, severity calls, and cross-team escalation.<\/li>\n<\/ul>\n<p>Two roles are easy to forget and costly to skip. <strong>Product\/Brand Marketing<\/strong> is the only role that can authoritatively say &quot;yes, that&#39;s the right number&quot; \u2014 they own the canon. And <strong>Technical SEO<\/strong> matters because a wrong answer often traces to a page the assistant simply couldn&#39;t crawl. If assistants can&#39;t read you, no content fix lands \u2014 crawlability is its own discipline that sits underneath every other fix.<\/p>\n<h2>The AI search RACI matrix<\/h2>\n<p><strong>Here is the core matrix.<\/strong> Read each row as a task and each cell as that role&#39;s job: <strong>A<\/strong> = accountable (one per row), <strong>R<\/strong> = responsible doer, <strong>C<\/strong> = consulted, <strong>I<\/strong> = informed.<\/p>\n<table>\n<thead>\n<tr>\n<th>Task<\/th>\n<th>GEO\/AEO Lead<\/th>\n<th>Content Owner<\/th>\n<th>Tech SEO \/ Web<\/th>\n<th>PR &amp; Comms<\/th>\n<th>Product \/ Brand<\/th>\n<th>Data \/ Analytics<\/th>\n<th>Exec Sponsor<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Detect &amp; triage incident<\/td>\n<td><strong>A\/R<\/strong><\/td>\n<td>I<\/td>\n<td>C<\/td>\n<td>I<\/td>\n<td>I<\/td>\n<td>C<\/td>\n<td>I<\/td>\n<\/tr>\n<tr>\n<td>Confirm root cause<\/td>\n<td>C<\/td>\n<td>C<\/td>\n<td><strong>A\/R<\/strong><\/td>\n<td>I<\/td>\n<td>C<\/td>\n<td>R<\/td>\n<td>I<\/td>\n<\/tr>\n<tr>\n<td>Approve corrected source-of-truth<\/td>\n<td>R<\/td>\n<td>C<\/td>\n<td>I<\/td>\n<td>C<\/td>\n<td><strong>A<\/strong><\/td>\n<td>I<\/td>\n<td>I<\/td>\n<\/tr>\n<tr>\n<td>Update owned content &amp; on-site facts<\/td>\n<td>C<\/td>\n<td><strong>A\/R<\/strong><\/td>\n<td>R<\/td>\n<td>I<\/td>\n<td>C<\/td>\n<td>I<\/td>\n<td>I<\/td>\n<\/tr>\n<tr>\n<td>Repair off-site citations<\/td>\n<td>R<\/td>\n<td>C<\/td>\n<td>I<\/td>\n<td><strong>A<\/strong><\/td>\n<td>C<\/td>\n<td>I<\/td>\n<td>I<\/td>\n<\/tr>\n<tr>\n<td>Re-test &amp; validate the fix<\/td>\n<td><strong>A<\/strong><\/td>\n<td>C<\/td>\n<td>C<\/td>\n<td>I<\/td>\n<td>I<\/td>\n<td>R<\/td>\n<td>I<\/td>\n<\/tr>\n<tr>\n<td>Report business risk &amp; ROI<\/td>\n<td><strong>A\/R<\/strong><\/td>\n<td>I<\/td>\n<td>I<\/td>\n<td>C<\/td>\n<td>C<\/td>\n<td>R<\/td>\n<td>I<\/td>\n<\/tr>\n<tr>\n<td>Escalate sev-1 \/ approve budget<\/td>\n<td>R<\/td>\n<td>I<\/td>\n<td>I<\/td>\n<td>C<\/td>\n<td>C<\/td>\n<td>C<\/td>\n<td><strong>A<\/strong><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>The discipline in this table is the <strong>single A per row<\/strong>. Eight tasks, eight clear owners, no shared accountability. The GEO\/AEO Lead owns detection, validation, and reporting \u2014 the program-level jobs \u2014 but <em>cannot<\/em> unilaterally approve a brand claim or rewrite an analyst&#39;s press quote. Those accountabilities sit with Product and PR on purpose.<\/p>\n<p>The next three sections walk the matrix through each failure type from the title.<\/p>\n<h3>Wrong answers: who does what<\/h3>\n<p><strong>A wrong answer is a factual error an assistant states about your brand<\/strong> \u2014 incorrect pricing, a feature you don&#39;t have, a misattributed founder, an outdated positioning line. It is the highest-urgency failure because it actively misleads buyers.<\/p>\n<p>Routing, per the matrix: the <strong>GEO\/AEO Lead<\/strong> detects and reproduces it (Accountable). <strong>Tech SEO<\/strong> confirms root cause \u2014 often a stale page, a blocked crawler, or a schema gap. <strong>Product\/Brand Marketing<\/strong> approves the corrected fact (the canon). <strong>Content Owner<\/strong> updates owned pages. The Lead re-tests until the assistant repeats the right answer.<\/p>\n<p>The trap is skipping the approval step. A writer &quot;corrects&quot; the claim using their best guess, and now your site contradicts your sales deck. Approval is a 15-minute decision that prevents a month of drift. The end-to-end mechanics of finding the offending source and replacing it live in a dedicated <a href=\"https:\/\/maxaeo.ai\/blog\/fix-wrong-ai-answer-about-my-brand\">source-repair workflow for wrong AI answers<\/a>.<\/p>\n<h3>Bad citations: who does what<\/h3>\n<p><strong>A bad citation is when an assistant attributes a claim about you to the wrong, outdated, or low-quality source<\/strong> \u2014 or cites a competitor&#39;s page as the authority on your own product. It corrodes ai reputation management slowly, because each individual instance looks minor, which is why a <a href=\"https:\/\/maxaeo.ai\/blog\/ai-brand-reputation-management-how-to-detect-and-fix-wrong-ai-answers-about-your-company\">system for detecting and fixing wrong AI answers about your company<\/a> matters more than any single fix.<\/p>\n<p>Per the matrix, <strong>PR &amp; Comms is Accountable<\/strong> for repairing off-site citations, with the GEO\/AEO Lead responsible for the legwork and Content consulted. Why PR? Because the fix usually lives off your domain \u2014 getting a review platform updated, a directory corrected, or an outdated press mention refreshed. These are earned-source relationships, and earned sources are PR&#39;s home turf.<\/p>\n<p>Stale sources are the most common culprit. An assistant keeps citing a two-year-old comparison because nothing newer ranks. The remedy is publishing fresher, more citable material \u2014 a process worth systematizing, as in <a href=\"https:\/\/maxaeo.ai\/blog\/outdated-ai-citations\">finding, prioritizing, and fixing the stale sources behind outdated AI citations<\/a>.<\/p>\n<h3>Lost recommendations: who does what<\/h3>\n<p><strong>A lost recommendation is when an assistant builds a shortlist for your category and omits you<\/strong> \u2014 or drops you from a list you previously made. This is the quietest and most expensive failure, because nothing is &quot;wrong&quot;; you&#39;re simply absent.<\/p>\n<p>This failure type stresses the matrix differently. Detection sits with the GEO\/AEO Lead and <strong>Data\/Analytics<\/strong>, who tracks share-of-voice on category prompts via llm brand tracking. Root cause is rarely a single broken page \u2014 it&#39;s thin entity coverage, weak third-party validation, or competitors with stronger citation footprints.<\/p>\n<p>The fix is a build, not a repair: stronger brand facts, more earned mentions, clearer category positioning. That&#39;s a campaign, so the <strong>Executive Sponsor<\/strong> often gets pulled in to fund it. Getting recommended by ChatGPT is won over weeks, not in an incident ticket \u2014 but the RACI still routes who owns the work.<\/p>\n<h2>AI search incident severity tiers<\/h2>\n<p><strong>Severity decides how far up the matrix an incident escalates.<\/strong> Tie tiers to revenue and reputation exposure, not to how loud the Slack thread is. A simple three-tier scale keeps the Executive Sponsor row from firing on every minor drift.<\/p>\n<table>\n<thead>\n<tr>\n<th>Tier<\/th>\n<th>Trigger<\/th>\n<th>Response time<\/th>\n<th>Owner<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><strong>Sev-1<\/strong><\/td>\n<td>Wrong answer on a buying-stage prompt, or a compliance\/safety falsehood<\/td>\n<td>Same day<\/td>\n<td>Escalate to Executive Sponsor<\/td>\n<\/tr>\n<tr>\n<td><strong>Sev-2<\/strong><\/td>\n<td>Bad citation, or a wrong answer on a mid-funnel prompt; reputation drift<\/td>\n<td>This week<\/td>\n<td>GEO\/AEO Lead<\/td>\n<\/tr>\n<tr>\n<td><strong>Sev-3<\/strong><\/td>\n<td>Lost recommendation, slow share-of-voice erosion, single low-traffic prompt<\/td>\n<td>This sprint \/ backlog<\/td>\n<td>Data + Content, batched<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>The hard part is knowing which red flag is a Sev-1 versus noise that can wait. That judgment call \u2014 <a href=\"https:\/\/maxaeo.ai\/blog\/ai-search-alerts\">which AI visibility changes need same-day action<\/a> \u2014 is what keeps the matrix from either crying wolf or moving too slowly.<\/p>\n<h2>Worked example: a wrong answer hits a buying-stage prompt<\/h2>\n<p><strong>Here is the matrix in motion.<\/strong> Treat this as a representative incident \u2014 the shape repeats across brands that monitor AI search daily.<\/p>\n<p>A monitoring alert fires Tuesday morning: asked &quot;is [your product] SOC 2 compliant?&quot;, ChatGPT and Perplexity both answer <strong>&quot;no&quot;<\/strong> \u2014 when you&#39;ve been compliant for eight months. This is a Sev-1 wrong answer on a buying-stage prompt, the worst kind.<\/p>\n<p>The RACI runs the response:<\/p>\n<ol>\n<li><strong>Diagnose (GEO\/AEO Lead, A).<\/strong> The Lead reproduces it across four assistants \u2014 confirmed on ChatGPT and Perplexity, correct on Gemini. Not a fluke.<\/li>\n<li><strong>Root cause (Tech SEO, A).<\/strong> Your trust page exists but is gated behind a form and blocked in robots.txt. Assistants can&#39;t read it; they&#39;re inferring &quot;no&quot; from an old blog post.<\/li>\n<li><strong>Approve the fact (Product\/Brand, A).<\/strong> Product confirms the exact compliant language and effective date. Five minutes.<\/li>\n<li><strong>Update (Content + Tech SEO, R).<\/strong> Content publishes an ungated, crawlable trust summary with explicit compliance facts; Tech SEO unblocks the path and adds schema.<\/li>\n<li><strong>Repair off-site (PR, A).<\/strong> PR updates the security listing on the two review platforms the assistants cite.<\/li>\n<li><strong>Validate (GEO\/AEO Lead, A).<\/strong> Within two crawl cycles, all four assistants answer &quot;yes.&quot; The Lead confirms and closes.<\/li>\n<li><strong>Report (Data, R \/ Lead, A).<\/strong> Data quantifies the window of exposure and the recovered mention rate for the leadership update.<\/li>\n<\/ol>\n<p>Total elapsed time with a RACI: <strong>under a week.<\/strong> Without one, that same incident often sits for a month while people debate whose problem it is. The matrix didn&#39;t fix the page \u2014 it removed the hesitation.<\/p>\n<h2>The mistake that breaks most AI search RACIs<\/h2>\n<p><strong>The single most common failure is confusing Responsible with Accountable.<\/strong> Teams assume the person doing the work also owns the outcome. In AI search, that quietly puts the busiest doer \u2014 usually the GEO\/AEO Lead \u2014 as accountable for <em>everything<\/em>, which means nothing has a true owner.<\/p>\n<p>Accountability is a <strong>decision right, not a workload<\/strong>. The person accountable for &quot;approve corrected source-of-truth&quot; is whoever can authoritatively declare the claim correct \u2014 Product or Brand, even if they write zero words. Conflate the two and you get the documented RACI failure mode: diluted ownership, finger-pointing, and fixes that stall at the approval gap.<\/p>\n<p>Three rules keep an AI search RACI honest:<\/p>\n<ul>\n<li><strong>One A per row, always.<\/strong> Two accountables means zero accountability.<\/li>\n<li><strong>A \u2260 the doer.<\/strong> Assign A by who decides, not who types.<\/li>\n<li><strong>Match A to severity.<\/strong> Sev-1 incidents escalate to the Executive Sponsor, who owns the budget call.<\/li>\n<\/ul>\n<p>This is where an AI search RACI connects to broader governance: deciding, as an organization, that marketing owns brand accuracy inside AI answers in the first place. That ownership question is the foundation the matrix sits on \u2014 explored in <a href=\"https:\/\/maxaeo.ai\/blog\/ai-search-governance\">how marketing teams should own brand accuracy in AI answers<\/a>.<\/p>\n<h2>How to roll out your AI search RACI in a week<\/h2>\n<p><strong>You can stand up a working AI search RACI in five steps.<\/strong> Don&#39;t over-engineer it; a one-page matrix your team actually uses beats a perfect document nobody opens.<\/p>\n<ol>\n<li><strong>Name the seven roles<\/strong> against real people (or merge them for a small team \u2014 one person can hold three columns).<\/li>\n<li><strong>Adopt the eight-row matrix above<\/strong> as your starting draft; edit cells to fit your org&#39;s authority lines.<\/li>\n<li><strong>Set severity tiers<\/strong> using the three-tier scale above, so the matrix knows when to escalate to the Executive Sponsor \u2014 tie tiers to revenue and reputation, not gut feel.<\/li>\n<li><strong>Wire it to your monitoring.<\/strong> The RACI is only as fast as your detection. Daily ai search monitoring across ChatGPT, Gemini, Perplexity, Claude, Copilot, Grok, and AI Overviews is what triggers row one.<\/li>\n<li><strong>Review it monthly<\/strong> against closed incidents. If a fix stalled, find which cell was ambiguous and tighten it.<\/li>\n<\/ol>\n<p>The reporting row is what earns the program its budget. Connecting incident volume and recovery to pipeline risk turns &quot;AI is misbehaving&quot; into a defensible number \u2014 the method is laid out in <a href=\"https:\/\/maxaeo.ai\/blog\/ai-search-monitoring-roi\">connecting mention rate, citations, and pipeline risk into monitoring ROI<\/a>.<\/p>\n<h2>Frequently asked questions<\/h2>\n<h3>What does RACI stand for in AI search?<\/h3>\n<p>RACI stands for <strong>Responsible, Accountable, Consulted, Informed<\/strong> \u2014 the same four roles used in project management, applied here to AI search incidents. Responsible does the work, Accountable owns the outcome (one person only), Consulted gives input, and Informed is kept in the loop. The novelty is the tasks: detecting wrong answers, approving corrected facts, repairing ai citations, and reporting business risk.<\/p>\n<h3>Who is accountable when ChatGPT says something wrong about my brand?<\/h3>\n<p>In a well-built AI search RACI, accountability splits by job. The <strong>GEO\/AEO Lead<\/strong> is accountable for detecting and validating the fix, <strong>Product or Brand Marketing<\/strong> is accountable for approving the correct fact, and <strong>PR<\/strong> is accountable for repairing off-site sources. No single person owns all of it \u2014 that&#39;s the whole point of the matrix.<\/p>\n<h3>How is an AI search RACI different from a normal marketing RACI?<\/h3>\n<p>A normal marketing RACI maps planned work \u2014 campaigns, launches, approvals. An AI search RACI maps <strong>reactive incidents<\/strong> that cross content, PR, technical, and data desks at once. The rows are failure types unique to answer engine optimization and generative engine optimization, and the urgency model is closer to incident response than campaign planning.<\/p>\n<h3>Do small teams need an AI search RACI?<\/h3>\n<p>Yes, but a compressed one. A three-person team can merge columns \u2014 one person holds GEO Lead, Content, and Data \u2014 and still benefit from the <strong>single-accountable-owner<\/strong> rule per task. The matrix prevents the small-team failure mode where everyone sees the alert and assumes a teammate is handling it.<\/p>\n<h3>What tools do you need to run an AI search RACI?<\/h3>\n<p>At minimum, a daily ai visibility tool that tracks mentions, citations, and recommendations across the major assistants, plus a shared matrix document and an alerting channel. The tracker triggers row one of the matrix; without reliable llm brand tracking, the RACI has nothing to respond to and incidents are caught by accident instead of by design.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>An AI search RACI assigns who diagnoses, approves, fixes, and reports every wrong answer, bad citation, and lost AI recommendation. Get the full role-by-role matrix.<\/p>\n","protected":false},"author":1,"featured_media":779,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-780","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/posts\/780","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/comments?post=780"}],"version-history":[{"count":0,"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/posts\/780\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/media\/779"}],"wp:attachment":[{"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/media?parent=780"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/categories?post=780"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/tags?post=780"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}