{"id":32572,"date":"2026-06-05T21:24:57","date_gmt":"2026-06-05T15:54:57","guid":{"rendered":"https:\/\/www.aicerts.ai\/news\/"},"modified":"2026-06-05T21:25:00","modified_gmt":"2026-06-05T15:55:00","slug":"anthropics-warning-spotlights-ai-safety-risks","status":"publish","type":"news","link":"https:\/\/www.aicerts.ai\/news\/anthropics-warning-spotlights-ai-safety-risks\/","title":{"rendered":"Anthropic\u2019s Warning Spotlights AI Safety Risks"},"content":{"rendered":"\n<p>Moreover, the lab is urging peers to agree on a verifiable brake before runaway feedback loops appear. Such a mechanism would pause training if telemetry shows accelerating, uncontrolled capability gains. This article unpacks the numbers, evaluates <strong>governance<\/strong> proposals, and outlines pragmatic steps for business leaders. Meanwhile, professionals seeking deeper expertise can pursue the <a href=\"https:\/\/www.aicerts.ai\/certifications\/business\/ai-researcher\">AI+ Researcher\u2122<\/a> certification.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" src=\"https:\/\/aicertswpcdn.blob.core.windows.net\/newsportal\/2026\/06\/ai-infrastructure-oversight.jpg\" alt=\"Data center infrastructure highlighting AI Safety Risks and model oversight\"\/><figcaption class=\"wp-element-caption\">Infrastructure and oversight go hand in hand as AI systems scale.<\/figcaption><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\">Anthropic Raises Red Flags<\/h2>\n\n\n\n<p>Public data reveal startling automation inside Anthropic\u2019s software pipeline. More than 80% of production code lines merged in May came from Claude\u2019s suggestions. Furthermore, engineers merged eight times more code per quarter versus 2024 baselines, thanks to model assistance.<\/p>\n\n\n\n<p>These metrics support the thesis that <strong>recursive self-improvement<\/strong> is already underway at a narrow scale. Consequently, management worries that future iterations could iterate on architecture, data, and evaluation without direct supervision. Jack Clark estimates a sixty percent chance of full self-training systems by 2028.<\/p>\n\n\n\n<p>Nevertheless, researchers stress the outcome is not inevitable if verification infrastructure matures in time. They argue transparent telemetry and strong <strong>alignment<\/strong> incentives can slow dangerous accelerations.<\/p>\n\n\n\n<p>Anthropic\u2019s internal figures show impressive but risky momentum. Verification gaps keep <strong>AI Safety Risks<\/strong> firmly on the table. Next, we decode how <strong>recursive self-improvement<\/strong> actually works.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">How Recursive Self-Improvement Works<\/h2>\n\n\n\n<p><strong>Recursive self-improvement<\/strong> describes an AI cycle that designs, tests, and trains a superior successor. Each generation can then repeat the loop, potentially creating exponential capability growth. Moreover, the process may compress research timelines from years to weeks, depending on compute availability.<\/p>\n\n\n\n<p>Anthropic models already optimize loss functions, training parameters, and even deployment scripts with minimal human nudges. In contrast, earlier systems required extensive manual tuning to integrate discoveries. Consequently, observers classify Claude and Mythos as early <strong>frontier models<\/strong> experimenting with partial autonomy.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Claude posted a 76% success rate on Anthropic\u2019s hardest open-ended coding tasks.<\/li>\n\n\n\n<li>Mythos achieved a 52\u00d7 kernel optimization speedup over baseline experiments.<\/li>\n\n\n\n<li>Some agent runs lasted 16 hours without failure, measured by external researchers.<\/li>\n\n\n\n<li>Engineering throughput rose eightfold per quarter versus 2024 benchmarks.<\/li>\n<\/ul>\n\n\n\n<p>These datapoints illustrate the technical plausibility of a self-improving loop. Yet unknown feedback dynamics sustain significant <strong>AI Safety Risks<\/strong>. We now examine what this means for larger <strong>frontier models<\/strong>.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Implications For Frontier Models<\/h2>\n\n\n\n<p><strong>Frontier models<\/strong> combine multimodal reasoning, long-term planning, and code generation at scale. Therefore, even modest <strong>recursive self-improvement<\/strong> could give these systems strategic autonomy. Regulators fear a single lab could control an accelerating capability curve before safeguards mature.<\/p>\n\n\n\n<p>Moreover, financial authorities worry about Mythos-class exploits that threaten banking infrastructure and cross-border payments. <strong>Alignment<\/strong> failures in such domains translate directly into systemic economic shocks. Consequently, the company and peers briefed central banks on potential cyber cascading effects.<\/p>\n\n\n\n<p>In contrast, cautious development promises societal benefits like rapid drug discovery and climate modeling. However, realizing those gains demands stringent <strong>governance<\/strong> across labs and supply chains.<\/p>\n\n\n\n<p>Frontier models magnify both scale and stakes. Unchecked growth will exacerbate <strong>AI Safety Risks<\/strong> for every sector. The following section explores the proposed safety brake mechanism.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Proposed Safety Brake Mechanism<\/h2>\n\n\n\n<p>The institute recommends a coordinated, verifiable pause protocol that activates under predefined risk thresholds. Moreover, participation would require multiple frontier labs to cryptographically prove compliance. This shared brake aims to prevent unilateral pauses that could backfire by shifting incentives abroad.<\/p>\n\n\n\n<p>Initially, telemetry standards must detect signs of self-directed capability jumps, such as automated architecture search. Furthermore, trusted auditors must certify logs to satisfy government oversight. Legal interoperability across jurisdictions will anchor long-term <strong>governance<\/strong> stability.<\/p>\n\n\n\n<p>Nevertheless, building such infrastructure requires incentives aligned with commercial timelines. Developers worry that delays could cede market share to less cooperative actors. Therefore, policymakers contemplate liability shields and procurement carrots to encourage adoption.<\/p>\n\n\n\n<p>A verifiable brake could mitigate runaway trajectories if global buy-in materializes. Implementation hurdles keep <strong>AI Safety Risks<\/strong> unresolved for now. Consequently, regulatory and market reactions deserve close scrutiny.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Regulatory And Market Responses<\/h2>\n\n\n\n<p>Regulators have already moved from observation to engagement. The Bank of England flagged Mythos capabilities as a direct cyber risk to liquidity providers. Meanwhile, the Financial Stability Board requested regular briefings on <strong>frontier models<\/strong> and audit trails.<\/p>\n\n\n\n<p>Across the Atlantic, US senators floated export controls tied to <strong>alignment<\/strong> benchmarks and transparency obligations. Moreover, venture capital firms are stress-testing portfolios against <strong>governance<\/strong> non-compliance scenarios. Insurance carriers likewise explore new risk classes for self-directed systems.<\/p>\n\n\n\n<p>Industry groups support harmonized standards but resist rules that privilege incumbents. Nevertheless, consumer sentiment increasingly rewards companies that disclose <strong>AI Safety Risks<\/strong> proactively. Consequently, enterprises seek certified talent to navigate shifting requirements. Professionals can strengthen credibility via the <a href=\"https:\/\/www.aicerts.ai\/certifications\/business\/ai-researcher\">AI+ Researcher\u2122<\/a> certification.<\/p>\n\n\n\n<p>Regulatory momentum is growing but remains fragmented. Harmonization gaps sustain persistent <strong>AI Safety Risks<\/strong>. Leaders must translate these signals into actionable roadmaps.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Action Steps For Leaders<\/h2>\n\n\n\n<p>Executives should prioritize internal telemetry that tracks model contributions, code merges, and compute use. Furthermore, board committees need quarterly reviews of <strong>alignment<\/strong> metrics and breach simulations. Cross-industry alliances can pool resources for open <strong>governance<\/strong> tooling.<\/p>\n\n\n\n<p>Firms can adopt phased build checkpoints that trigger audits before major capability upgrades. Moreover, scenario planning should include supply-chain shocks, reputational fallout, and legal exposure. Dedicated red-team exercises will surface hidden <strong>AI Safety Risks<\/strong> early.<\/p>\n\n\n\n<p>Meanwhile, talent development remains essential. Consequently, managers should sponsor staff pursuing the <a href=\"https:\/\/www.aicerts.ai\/certifications\/business\/ai-researcher\">AI+ Researcher\u2122<\/a> credential. Certified personnel can interface effectively with regulators and auditors.<\/p>\n\n\n\n<p>Structured processes, skilled teams, and shared standards reduce uncertainty. Persistent vigilance still matters because <strong>AI Safety Risks<\/strong> will evolve. Let us recap the core insights.<\/p>\n\n\n\n<p>The latest warning underscores an inflection point for advanced AI development. Recursive cycles, <strong>frontier models<\/strong>, and shaky policy now intersect with real economic stakes. Therefore, leaders must install telemetry, risk metrics, and verifiable brakes before acceleration outpaces oversight. Meanwhile, collaborative policy frameworks can convert competitive tension into shared safety dividends. Professionals should continually upgrade knowledge through recognized credentials like the <a href=\"https:\/\/www.aicerts.ai\/certifications\/business\/ai-researcher\">AI+ Researcher\u2122<\/a> program. Taking these steps today will mitigate tomorrow\u2019s <strong>AI Safety Risks<\/strong> while unlocking responsible innovation.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Critical questions are mounting after Anthropic issued its starkest warning yet. The company argues that advanced AI may soon redesign itself faster than humans can track. Consequently, executives and regulators face new AI Safety Risks that extend far beyond hypothetical debates. Anthropic\u2019s essay, \u201cWhen AI builds itself\u201d, highlights early signals of recursive self-improvement.<\/p>\n","protected":false},"featured_media":32569,"parent":0,"comment_status":"open","ping_status":"closed","template":"","meta":{"_acf_changed":false,"_yoast_wpseo_focuskw":"AI Safety Risks","_yoast_wpseo_title":"","_yoast_wpseo_metadesc":"Anthropic's new warning on AI Safety Risks explores recursive self-improvement, frontier models, and governance steps leaders must grasp now.","_yoast_wpseo_canonical":""},"tags":[334,255,110,1571,69,8,15,55,43292],"news_category":[4,3,6],"communities":[],"class_list":["post-32572","news","type-news","status-publish","has-post-thumbnail","hentry","tag-ai-certifications","tag-ai-certs","tag-ai-innovation","tag-ai-platform","tag-ai-tools","tag-artificial-intelligence","tag-generative-ai","tag-productivity-tools","tag-recursive-ai","news_category-ai","news_category-business","news_category-machine-learning"],"acf":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.6 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Anthropic\u2019s Warning Spotlights AI Safety Risks - AI CERTs News<\/title>\n<meta name=\"description\" content=\"Anthropic&#039;s new warning on AI Safety Risks explores recursive self-improvement, frontier models, and governance steps leaders must grasp now.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.aicerts.ai\/news\/anthropics-warning-spotlights-ai-safety-risks\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Anthropic\u2019s Warning Spotlights AI Safety Risks - AI CERTs News\" \/>\n<meta property=\"og:description\" content=\"Anthropic&#039;s new warning on AI Safety Risks explores recursive self-improvement, frontier models, and governance steps leaders must grasp now.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.aicerts.ai\/news\/anthropics-warning-spotlights-ai-safety-risks\/\" \/>\n<meta property=\"og:site_name\" content=\"AI CERTs News\" \/>\n<meta property=\"article:modified_time\" content=\"2026-06-05T15:55:00+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/aicertswpcdn.blob.core.windows.net\/newsportal\/2026\/06\/policy-review-meeting-6a22a315027b3.jpg\" \/>\n\t<meta property=\"og:image:width\" content=\"1024\" \/>\n\t<meta property=\"og:image:height\" content=\"576\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/jpeg\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data1\" content=\"5 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/anthropics-warning-spotlights-ai-safety-risks\\\/\",\"url\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/anthropics-warning-spotlights-ai-safety-risks\\\/\",\"name\":\"Anthropic\u2019s Warning Spotlights AI Safety Risks - AI CERTs News\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/anthropics-warning-spotlights-ai-safety-risks\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/anthropics-warning-spotlights-ai-safety-risks\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/aicertswpcdn.blob.core.windows.net\\\/newsportal\\\/2026\\\/06\\\/policy-review-meeting-6a22a315027b3.jpg\",\"datePublished\":\"2026-06-05T15:54:57+00:00\",\"dateModified\":\"2026-06-05T15:55:00+00:00\",\"description\":\"Anthropic's new warning on AI Safety Risks explores recursive self-improvement, frontier models, and governance steps leaders must grasp now.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/anthropics-warning-spotlights-ai-safety-risks\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/anthropics-warning-spotlights-ai-safety-risks\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/anthropics-warning-spotlights-ai-safety-risks\\\/#primaryimage\",\"url\":\"https:\\\/\\\/aicertswpcdn.blob.core.windows.net\\\/newsportal\\\/2026\\\/06\\\/policy-review-meeting-6a22a315027b3.jpg\",\"contentUrl\":\"https:\\\/\\\/aicertswpcdn.blob.core.windows.net\\\/newsportal\\\/2026\\\/06\\\/policy-review-meeting-6a22a315027b3.jpg\",\"width\":1024,\"height\":576,\"caption\":\"Decision-makers reviewing the latest guidance on AI safety risks.\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/anthropics-warning-spotlights-ai-safety-risks\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"News\",\"item\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/news\\\/\"},{\"@type\":\"ListItem\",\"position\":3,\"name\":\"Anthropic\u2019s Warning Spotlights AI Safety Risks\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/#website\",\"url\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/\",\"name\":\"Aicerts News\",\"description\":\"\",\"publisher\":{\"@id\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/#organization\",\"name\":\"Aicerts News\",\"url\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/wp-content\\\/uploads\\\/2024\\\/09\\\/news_logo.svg\",\"contentUrl\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/wp-content\\\/uploads\\\/2024\\\/09\\\/news_logo.svg\",\"width\":1,\"height\":1,\"caption\":\"Aicerts News\"},\"image\":{\"@id\":\"https:\\\/\\\/www.aicerts.ai\\\/news\\\/#\\\/schema\\\/logo\\\/image\\\/\"}}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Anthropic\u2019s Warning Spotlights AI Safety Risks - AI CERTs News","description":"Anthropic's new warning on AI Safety Risks explores recursive self-improvement, frontier models, and governance steps leaders must grasp now.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.aicerts.ai\/news\/anthropics-warning-spotlights-ai-safety-risks\/","og_locale":"en_US","og_type":"article","og_title":"Anthropic\u2019s Warning Spotlights AI Safety Risks - AI CERTs News","og_description":"Anthropic's new warning on AI Safety Risks explores recursive self-improvement, frontier models, and governance steps leaders must grasp now.","og_url":"https:\/\/www.aicerts.ai\/news\/anthropics-warning-spotlights-ai-safety-risks\/","og_site_name":"AI CERTs News","article_modified_time":"2026-06-05T15:55:00+00:00","og_image":[{"width":1024,"height":576,"url":"https:\/\/aicertswpcdn.blob.core.windows.net\/newsportal\/2026\/06\/policy-review-meeting-6a22a315027b3.jpg","type":"image\/jpeg"}],"twitter_card":"summary_large_image","twitter_misc":{"Est. reading time":"5 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.aicerts.ai\/news\/anthropics-warning-spotlights-ai-safety-risks\/","url":"https:\/\/www.aicerts.ai\/news\/anthropics-warning-spotlights-ai-safety-risks\/","name":"Anthropic\u2019s Warning Spotlights AI Safety Risks - AI CERTs News","isPartOf":{"@id":"https:\/\/www.aicerts.ai\/news\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.aicerts.ai\/news\/anthropics-warning-spotlights-ai-safety-risks\/#primaryimage"},"image":{"@id":"https:\/\/www.aicerts.ai\/news\/anthropics-warning-spotlights-ai-safety-risks\/#primaryimage"},"thumbnailUrl":"https:\/\/aicertswpcdn.blob.core.windows.net\/newsportal\/2026\/06\/policy-review-meeting-6a22a315027b3.jpg","datePublished":"2026-06-05T15:54:57+00:00","dateModified":"2026-06-05T15:55:00+00:00","description":"Anthropic's new warning on AI Safety Risks explores recursive self-improvement, frontier models, and governance steps leaders must grasp now.","breadcrumb":{"@id":"https:\/\/www.aicerts.ai\/news\/anthropics-warning-spotlights-ai-safety-risks\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.aicerts.ai\/news\/anthropics-warning-spotlights-ai-safety-risks\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.aicerts.ai\/news\/anthropics-warning-spotlights-ai-safety-risks\/#primaryimage","url":"https:\/\/aicertswpcdn.blob.core.windows.net\/newsportal\/2026\/06\/policy-review-meeting-6a22a315027b3.jpg","contentUrl":"https:\/\/aicertswpcdn.blob.core.windows.net\/newsportal\/2026\/06\/policy-review-meeting-6a22a315027b3.jpg","width":1024,"height":576,"caption":"Decision-makers reviewing the latest guidance on AI safety risks."},{"@type":"BreadcrumbList","@id":"https:\/\/www.aicerts.ai\/news\/anthropics-warning-spotlights-ai-safety-risks\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.aicerts.ai\/news\/"},{"@type":"ListItem","position":2,"name":"News","item":"https:\/\/www.aicerts.ai\/news\/news\/"},{"@type":"ListItem","position":3,"name":"Anthropic\u2019s Warning Spotlights AI Safety Risks"}]},{"@type":"WebSite","@id":"https:\/\/www.aicerts.ai\/news\/#website","url":"https:\/\/www.aicerts.ai\/news\/","name":"Aicerts News","description":"","publisher":{"@id":"https:\/\/www.aicerts.ai\/news\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.aicerts.ai\/news\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/www.aicerts.ai\/news\/#organization","name":"Aicerts News","url":"https:\/\/www.aicerts.ai\/news\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.aicerts.ai\/news\/#\/schema\/logo\/image\/","url":"https:\/\/www.aicerts.ai\/news\/wp-content\/uploads\/2024\/09\/news_logo.svg","contentUrl":"https:\/\/www.aicerts.ai\/news\/wp-content\/uploads\/2024\/09\/news_logo.svg","width":1,"height":1,"caption":"Aicerts News"},"image":{"@id":"https:\/\/www.aicerts.ai\/news\/#\/schema\/logo\/image\/"}}]}},"_links":{"self":[{"href":"https:\/\/www.aicerts.ai\/news\/wp-json\/wp\/v2\/news\/32572","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.aicerts.ai\/news\/wp-json\/wp\/v2\/news"}],"about":[{"href":"https:\/\/www.aicerts.ai\/news\/wp-json\/wp\/v2\/types\/news"}],"replies":[{"embeddable":true,"href":"https:\/\/www.aicerts.ai\/news\/wp-json\/wp\/v2\/comments?post=32572"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.aicerts.ai\/news\/wp-json\/wp\/v2\/media\/32569"}],"wp:attachment":[{"href":"https:\/\/www.aicerts.ai\/news\/wp-json\/wp\/v2\/media?parent=32572"}],"wp:term":[{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.aicerts.ai\/news\/wp-json\/wp\/v2\/tags?post=32572"},{"taxonomy":"news_category","embeddable":true,"href":"https:\/\/www.aicerts.ai\/news\/wp-json\/wp\/v2\/news_category?post=32572"},{"taxonomy":"communities","embeddable":true,"href":"https:\/\/www.aicerts.ai\/news\/wp-json\/wp\/v2\/communities?post=32572"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}