{"id":209879,"date":"2026-05-06T12:02:00","date_gmt":"2026-05-06T16:02:00","guid":{"rendered":"https:\/\/testing.news-you-need.com\/index.php\/2026\/05\/06\/cybersecurity-agencies-issue-new-guidance-for-adopting-agentic-ai\/"},"modified":"2026-05-06T17:40:16","modified_gmt":"2026-05-06T21:40:16","slug":"cybersecurity-agencies-issue-new-guidance-for-adopting-agentic-ai","status":"publish","type":"post","link":"https:\/\/testing.news-you-need.com\/index.php\/2026\/05\/06\/cybersecurity-agencies-issue-new-guidance-for-adopting-agentic-ai\/","title":{"rendered":"Cybersecurity agencies issue new guidance for adopting agentic AI"},"content":{"rendered":"<p><a href=\"https:\/\/www.biometricupdate.com\/202605\/cybersecurity-agencies-issue-new-guidance-for-adopting-agentic-ai\">Cybersecurity agencies issue new guidance for adopting agentic AI<\/a><\/p>\n<p><a href=\"https:\/\/www.biometricupdate.com\/202605\/cybersecurity-agencies-issue-new-guidance-for-adopting-agentic-ai\">https:\/\/www.biometricupdate.com\/202605\/cybersecurity-agencies-issue-new-guidance-for-adopting-agentic-ai<\/a><\/p>\n<p>Publish Date: <a href=\"publish_date]\">2026-05-06 12:02:00<\/a><\/p>\n<p>Source Domain: <a href=\"www.biometricupdate.com\">www.biometricupdate.com<\/a><\/p>\n<p>Author: <a href=\"\"><\/a><\/p>\n<p> Using an unordered list, summarize the following article with between 4 and 8 key points. <\/p>\n<p>\t\t\t\tNew guidance developed by a group of national cybersecurity agencies aims to lay out \u201ckey cyber security challenges and risks associated with the introduction of agentic AI into IT environments, as well as best practices for securing agentic AI systems.\u201d<br \/>\nAuthoring agencies include the Australian Signals Directorate\u2019s Australian Cyber Security Centre (ASD\u2019s ACSC), the United States Cybersecurity and Infrastructure Security Agency (CISA) and National Security Agency (NSA), the Canadian Centre for Cyber Security (Cyber Centre), the New Zealand National Cyber Security Centre (NCSC-NZ) and the United Kingdom National Cyber Security Centre (NCSC-UK).<\/p>\n<p>In their report on \u201ccareful adoption of agentic AI services,\u201d which focuses on large language model (LLM)-based agentic AI systems, the agencies together \u201cstrongly recommend aligning agentic AI risks and mitigation strategies with your organisation\u2019s existing security model and risk posture.\u201d Adopt agentic AI with a security-first mindset, and do not grant it broad or unrestricted access to sensitive data or critical systems; ideally, agents should only be used for low-risk and non-sensitive tasks.<br \/>\nGuidance identifies best practices for design, development, deployment<br \/>\nThe guidance lists the security risks that agentic AI poses, and lays out best practices for securing agentic systems.<br \/>\n\u201cPrivilege escalation, emergent behaviours, structural dependencies and accountability gaps can interact in unpredictable ways,\u201d it says. \u201cAgentic AI developers, vendors and operators should implement a layered defence and strict access controls to reduce the likelihood of compromise.\u201d Recommandations cover practical design, development and deployment of AI agents, which range from oversight mechanisms to layered defenses to comprehensive testing and evaluation.<br \/>\nAs observers of the biometrics sector will know, governance is a key concern for deployment of AI agents. \u201cAutonomous actions by agentic AI systems introduce new risks, requiring updated governance policies and continuous runtime authentication with centralized policy decision points for each action.\u201d<br \/>\n\u201cOngoing strict privilege management of AI agents is key to long-term security. Lapses here can change the impact of a buggy agent from minor to catastrophic.\u201d<br \/>\nFuture-proofing requires collaboration, cool heads<br \/>\nAgentic AI continues to develop, and responsible use is a long game. \u201cOrganizations must anticipate and address the new risks these systems introduce,\u201d the report says. \u201cWhile industry and academia are developing practices to secure agentic AI, the field is still evolving and requires continued research and the practical implementation of agent security to address emerging challenges.\u201d<br \/>\nThe authoring agencies recommend security practitioners and researchers expand threat intelligence through collaboration, develop robust and agent-specific evaluations, and \u201cuse system-theoretic approaches to analyse agentic AI systems and identify appropriate security measures.\u201d<br \/>\nAnd, overall, organizations should proceed with caution. \u201cDeploy agentic AI incrementally, beginning with clearly defined low\u2011risk tasks and continuously assessing it against evolving threat models. Strong governance, explicit accountability, rigorous monitoring and human oversight are not optional safeguards but essential prerequisites.\u201d<br \/>\n\u201cUntil security practices, evaluation methods and standards mature, organisations should assume that agentic AI systems may behave unexpectedly and plan deployments accordingly, prioritizing resilience, reversibility and risk containment over efficiency gains.\u201d<br \/>\nIn other words, say some of the world\u2019s foremost cybersecurity agencies, take a breather before releasing a horde of autonomous agents into your enterprise IT system. Agentic AI has been evangelized as the future of the moment. But the risk is too high to rush in.<\/p>\n<p>\t\t\t\tArticle Topics<br \/>\n\t\t\t\tAI agents \u00a0|\u00a0 CISA \u00a0|\u00a0 cybersecurity<\/p>\n<p>\t\t\t\t\tLatest Biometrics News<\/p>\n<p>\t\t\t\t\t\u00a0\t\t\t\t\t\t\t\t<\/p>\n<p>\t\t\t\t\t\t\t\tMay 6, 2026, 5:10 pm EDT<br \/>\n\t\t\t\t\t\t\t\t\t\t\tBy Harry Varatharasan, Chief Product Officer, ComplyCube The UK government\u02bcs digital identity consultation is closing, and for most commentators, this\u2026<\/p>\n<p>\t\t\t\t\t\u00a0\t\t\t\t\t\t\t\t<\/p>\n<p>\t\t\t\t\t\t\t\tMay 6, 2026, 4:44 pm EDT<br \/>\n\t\t\t\t\t\t\t\t\t\t\tLondon-based global identity verification and fraud prevention provider Shufti has passed a Level 3 evaluation of its biometric Presentation Attack\u2026<\/p>\n<p>\t\t\t\t\t\u00a0\t\t\t\t\t\t\t\t<\/p>\n<p>\t\t\t\t\t\t\t\tMay 6, 2026, 4:40 pm EDT<br \/>\n\t\t\t\t\t\t\t\t\t\t\tVoting is open for approval of a draft specification to extend OpenID Connect to cover new features for requesting and\u2026<\/p>\n<p>\t\t\t\t\t\u00a0\t\t\t\t\t\t\t\t<\/p>\n<p>\t\t\t\t\t\t\t\tMay 6, 2026, 4:24 pm EDT<br \/>\n\t\t\t\t\t\t\t\t\t\t\tCrowds, chaos and cranky travelers: The EU\u2019s biometric border management scheme, the Entry-Exit System (EES), continues to fill headlines as\u2026<\/p>\n<p>\t\t\t\t\t\u00a0\t\t\t\t\t\t\t\t<\/p>\n<p>\t\t\t\t\t\t\t\tMay 6, 2026, 2:52 pm EDT<br \/>\n\t\t\t\t\t\t\t\t\t\t\tThe Home Office is hosting a preliminary market engagement event to engage with potential suppliers for two not-yet-guaranteed future procurements\u2026<\/p>\n<p>\t\t\t\t\t\u00a0\t\t\t\t\t\t\t\t<\/p>\n<p>\t\t\t\t\t\t\t\tMay 6, 2026, 11:30 am EDT<br \/>\n\t\t\t\t\t\t\t\t\t\t\tMeta says it has begun using AI to detect and remove users under 13 from its platforms, and to automatically\u2026<\/p>\n<p><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Cybersecurity agencies issue new guidance for adopting agentic AI https:\/\/www.biometricupdate.com\/202605\/cybersecurity-agencies-issue-new-guidance-for-adopting-agentic-ai Publish Date: 2026-05-06 12:02:00 Source&#8230;<\/p>\n","protected":false},"author":1,"featured_media":209880,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"fifu_image_url":"https:\/\/d1sr9z1pdl3mb7.cloudfront.net\/wp-content\/uploads\/2026\/01\/15204113\/kya-scaled.jpg","fifu_image_alt":"","footnotes":""},"categories":[15],"tags":[26,24,18,17],"class_list":["post-209879","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-cybersecurity","tag-ai","tag-cybersecurity","tag-large-language-model","tag-llm"],"_links":{"self":[{"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/posts\/209879"}],"collection":[{"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/comments?post=209879"}],"version-history":[{"count":1,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/posts\/209879\/revisions"}],"predecessor-version":[{"id":209881,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/posts\/209879\/revisions\/209881"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/media\/209880"}],"wp:attachment":[{"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/media?parent=209879"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/categories?post=209879"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/testing.news-you-need.com\/index.php\/wp-json\/wp\/v2\/tags?post=209879"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}