{"id":132018,"date":"2026-08-06T19:41:11","date_gmt":"2026-08-06T19:41:11","guid":{"rendered":"https:\/\/www.europesays.com\/ai\/132018\/"},"modified":"2026-08-06T19:41:11","modified_gmt":"2026-08-06T19:41:11","slug":"ai-agents-need-speed-humans-simply-cant-match","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/ai\/132018\/","title":{"rendered":"AI Agents Need Speed Humans Simply Can&#8217;t Match"},"content":{"rendered":"<p>                    McKesson&#8217;s Arsh Arora on Building Guardrails for Autonomous AI<\/p>\n<p>                                                <a class=\"author-link\" href=\"https:\/\/www.bankinfosecurity.com\/authors\/michael-novinson-i-4923\" rel=\"nofollow noopener\" target=\"_blank\">Michael Novinson<\/a> (<a href=\"https:\/\/www.twitter.com\/MichaelNovinson\" rel=\"nofollow noopener\" target=\"_blank\">MichaelNovinson<\/a>)                                                    \u2022<br \/>\n                        August 6, 2026 \u00a0 \u00a0 <a href=\"https:\/\/www.bankinfosecurity.com\/ai-agents-need-speed-humans-simply-cant-match-a-32440#disqus_thread\" rel=\"nofollow noopener\" target=\"_blank\"><\/p>\n<p>                <img decoding=\"async\" src=\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/08\/ai-agents-need-speed-humans-simply-cant-match-image_large-6-a-32440.jpg\" alt=\"AI Agents Need Speed Humans Simply Can't Match\" class=\"img-responsive hidden\"\/><\/p>\n<p>Arsh Arora, cyber incident response and AI operations lead, McKesson<\/p>\n<p>Organizations racing to deploy artificial intelligence agents are often handing over broad access without needed guardrails.<\/p>\n<p>See Also: <a href=\"https:\/\/www.bankinfosecurity.com\/healthcare-leaders-are-rethinking-their-data-strategy-before-scaling-ai-a-32409?rf=RAM_SeeAlso\" rel=\"nofollow noopener\" target=\"_blank\">Why Healthcare Leaders Are Rethinking Their Data Strategy Before Scaling AI<\/a><\/p>\n<p>Arsh Arora, cyber incident response and AI operations lead at McKesson, pointed to an agent that was granted full production access to fix a bug and instead went &#8220;from savior to destroyer in nine seconds,&#8221; deleting the company&#8217;s database and backups &#8211; illustrating why human-speed oversight can&#8217;t keep pace with AI-speed threats.<\/p>\n<p>He said the fix isn&#8217;t avoiding AI agents but building narrowly scoped ones for specific tasks rather than one system with sweeping access, then validating their output under a &#8220;trust but verify&#8221; model. &#8220;We have compartmentalized the AI agents. We created agents that are for specific task, nothing else,&#8221; Arora said.<\/p>\n<p>In this video interview with ISMG at <a href=\"https:\/\/www.bankinfosecurity.com\/black-hat-c-372\" rel=\"nofollow noopener\" target=\"_blank\">Black Hat USA 2026<\/a>, Arora also discussed:<\/p>\n<p>The physical security risks of AI assistants reading calendars, emails and file access logs;<br \/>\nHow broad access to email and collaboration data can expose business and physical security risks;<br \/>\nWhy organizations should test small use cases before increasing agent autonomy. <\/p>\n<p>Arora has more than a decade of experience in threat intelligence, incident response and digital forensics. He has supported domestic and international investigations with the FBI and held security roles at Visa and Neustar. His expertise includes malware analysis, ransomware detection, open-source intelligence, bitcoin investigations, product management and financial crime and fraud mitigation.<\/p>\n<p>            <script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n","protected":false},"excerpt":{"rendered":"McKesson&#8217;s Arsh Arora on Building Guardrails for Autonomous AI Michael Novinson (MichaelNovinson) \u2022 August 6, 2026 \u00a0 \u00a0&hellip;\n","protected":false},"author":2,"featured_media":132019,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6],"tags":[405,1365,32086,1710,65828,7537,8174,65826,60670,65827,9578],"class_list":["post-132018","post","type-post","status-publish","format-standard","has-post-thumbnail","category-agentic-ai","tag-ai-agents","tag-ai-automation","tag-ai-guardrails","tag-ai-security","tag-arsh-arora","tag-artificial-intelligence-agents","tag-autonomous-ai","tag-cyber-incident-response","tag-human-oversight","tag-mckesson","tag-zero-trust"],"_links":{"self":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/132018","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/comments?post=132018"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/132018\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media\/132019"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media?parent=132018"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/categories?post=132018"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/tags?post=132018"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}