[{"data":1,"prerenderedAt":267},["ShallowReactive",2],{"blog-post-blog_en-nicht-menschliche-identitaeten-fuer-ki-agenten":3},{"id":4,"title":5,"body":6,"cover":251,"date":252,"description":253,"draft":254,"extension":255,"meta":256,"navigation":257,"path":258,"seo":259,"stem":260,"tags":261,"__hash__":266},"blog_en\u002Fen\u002Fblog\u002Fnicht-menschliche-identitaeten-fuer-ki-agenten.md","Non-Human Identities for AI Agents: IAM for Growing Teams",{"type":7,"value":8,"toc":246},"minimark",[9,13,18,26,29,57,60,64,67,70,208,223,226,230,233,242],[10,11,12],"p",{},"Non-human identities are moving from infrastructure detail to leadership topic because of AI agents. Once agents update tickets, read customer data, or trigger deployments, managing API keys is no longer enough. Teams need to clarify which digital identity an agent has, which permissions are attached to it, and who is accountable for those permissions.",[14,15,17],"h2",{"id":16},"what-non-human-identities-change","What Non-Human Identities Change",[10,19,20,21,25],{},"Non-human identities are ",[22,23,24],"strong",{},"service accounts, workload identities, API keys, bots, CI\u002FCD jobs, and AI agents"," that authenticate against systems. They are not new. What is new is the speed at which agents use these identities, combine them, and perform actions across multiple tools.",[10,27,28],{},"That turns IAM for automation into an architecture decision:",[30,31,32,39,45,51],"ul",{},[33,34,35,38],"li",{},[22,36,37],{},"Accountability:"," A log entry must show whether a human, an agent, or a backend service took the action.",[33,40,41,44],{},[22,42,43],{},"Least privilege:"," An agent needs specific capabilities, not inherited admin rights from a generic service account.",[33,46,47,50],{},[22,48,49],{},"Offboarding:"," Unused agents, tokens, and integrations must be removed just like former employees.",[33,52,53,56],{},[22,54,55],{},"Auditability:"," Decisions and tool calls must remain traceable, especially around customer data, payments, or production changes.",[10,58,59],{},"The OWASP Non-Human Identities Top 10 2025 names exactly these risks: improper offboarding, secret leakage, overprivileged identities, insecure authentication, and long-lived secrets. For growing companies, these are not edge cases for security teams. They become sources of operational risk and later compliance cost.",[14,61,63],{"id":62},"where-teams-should-start-before-adoption","Where Teams Should Start Before Adoption",[10,65,66],{},"The most common mistake is attaching AI agents to existing service accounts. It is fast, but it destroys the boundary between user action, automation, and system operation.",[10,68,69],{},"A solid starting point is a small identity model for agents:",[71,72,77],"pre",{"className":73,"code":74,"language":75,"meta":76,"style":76},"language-yaml shiki shiki-themes github-light github-dark","# Example: identity model for an internal release agent\nagent: release-assistant\nowner: platform-team\nidentity_type: workload_identity\nauth: oidc_short_lived_token\nallowed_actions: [\"read_ci_status\", \"create_release_note\", \"trigger_staging_deploy\"]\nforbidden_actions: [\"deploy_production\", \"change_permissions\", \"export_customer_data\"]\naudit_log: required\nreview_cycle: quarterly\n","yaml","",[78,79,80,89,104,115,126,137,163,186,197],"code",{"__ignoreMap":76},[81,82,85],"span",{"class":83,"line":84},"line",1,[81,86,88],{"class":87},"sJ8bj","# Example: identity model for an internal release agent\n",[81,90,92,96,100],{"class":83,"line":91},2,[81,93,95],{"class":94},"s9eBZ","agent",[81,97,99],{"class":98},"sVt8B",": ",[81,101,103],{"class":102},"sZZnC","release-assistant\n",[81,105,107,110,112],{"class":83,"line":106},3,[81,108,109],{"class":94},"owner",[81,111,99],{"class":98},[81,113,114],{"class":102},"platform-team\n",[81,116,118,121,123],{"class":83,"line":117},4,[81,119,120],{"class":94},"identity_type",[81,122,99],{"class":98},[81,124,125],{"class":102},"workload_identity\n",[81,127,129,132,134],{"class":83,"line":128},5,[81,130,131],{"class":94},"auth",[81,133,99],{"class":98},[81,135,136],{"class":102},"oidc_short_lived_token\n",[81,138,140,143,146,149,152,155,157,160],{"class":83,"line":139},6,[81,141,142],{"class":94},"allowed_actions",[81,144,145],{"class":98},": [",[81,147,148],{"class":102},"\"read_ci_status\"",[81,150,151],{"class":98},", ",[81,153,154],{"class":102},"\"create_release_note\"",[81,156,151],{"class":98},[81,158,159],{"class":102},"\"trigger_staging_deploy\"",[81,161,162],{"class":98},"]\n",[81,164,166,169,171,174,176,179,181,184],{"class":83,"line":165},7,[81,167,168],{"class":94},"forbidden_actions",[81,170,145],{"class":98},[81,172,173],{"class":102},"\"deploy_production\"",[81,175,151],{"class":98},[81,177,178],{"class":102},"\"change_permissions\"",[81,180,151],{"class":98},[81,182,183],{"class":102},"\"export_customer_data\"",[81,185,162],{"class":98},[81,187,189,192,194],{"class":83,"line":188},8,[81,190,191],{"class":94},"audit_log",[81,193,99],{"class":98},[81,195,196],{"class":102},"required\n",[81,198,200,203,205],{"class":83,"line":199},9,[81,201,202],{"class":94},"review_cycle",[81,204,99],{"class":98},[81,206,207],{"class":102},"quarterly\n",[10,209,210,211,214,215,218,219,222],{},"Before production use, leadership and engineering should make at least three decisions. ",[22,212,213],{},"Delegation:"," Does the agent act on behalf of a user or as its own technical actor? ",[22,216,217],{},"Permission design:"," Are roles modelled per agent, per workflow, or per team? ",[22,220,221],{},"Operations:"," Who rotates secrets, reviews logs, and removes identities that are no longer used?",[10,224,225],{},"Technically, short-lived OIDC tokens, secret managers, clear scopes, and separated environments help. Organisationally, ownership matters more: every non-human identity needs a business purpose, a technical owner, and a review date.",[14,227,229],{"id":228},"why-this-matters","Why This Matters",[10,231,232],{},"AI agents become economically interesting when they perform work inside existing systems. That is also where new paths to data, infrastructure, and business processes appear. Without clean non-human identities, these paths become hard to control: permissions grow with every experiment, logs lose meaning, and nobody feels responsible for old tokens.",[10,234,235,236,241],{},"Good identity architecture does not slow teams down. It reduces later friction. Teams can approve agents faster because permissions, boundaries, and responsibility are already defined. For founders, product leaders, and engineering managers, this is a cost and risk topic: fewer security exceptions, better traceability, and less technical debt in AI infrastructure. An ",[237,238,240],"a",{"href":239},"\u002Fen\u002F#packages","Architecture & AI Review"," can assess whether agent permissions, backend boundaries, and audit requirements fit together.",[243,244,245],"style",{},"html pre.shiki code .sJ8bj, html code.shiki .sJ8bj{--shiki-default:#6A737D;--shiki-dark:#6A737D}html pre.shiki code .s9eBZ, html code.shiki .s9eBZ{--shiki-default:#22863A;--shiki-dark:#85E89D}html pre.shiki code .sVt8B, html code.shiki .sVt8B{--shiki-default:#24292E;--shiki-dark:#E1E4E8}html pre.shiki code .sZZnC, html code.shiki .sZZnC{--shiki-default:#032F62;--shiki-dark:#9ECBFF}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"title":76,"searchDepth":91,"depth":91,"links":247},[248,249,250],{"id":16,"depth":91,"text":17},{"id":62,"depth":91,"text":63},{"id":228,"depth":91,"text":229},null,"2026-05-06","Non-human identities secure AI agents and automation. What teams should clarify around roles, secrets, audit logs, and ownership.",false,"md",{},true,"\u002Fen\u002Fblog\u002Fnicht-menschliche-identitaeten-fuer-ki-agenten",{"title":5,"description":253},"en\u002Fblog\u002Fnicht-menschliche-identitaeten-fuer-ki-agenten",[262,263,264,265],"AI","Cybersecurity","Software Architecture","Governance","c_uRx_Yw05gAUUkqDVBG2a28vTq_zc4nFc3VHvyEhX8",1780122462516]