Atera’s AI Shift: Gil Pekelman on Accountability and Risk in Autonomous IT for MSPs

The episode highlights a structural shift from automation that suggests actions to automation that executes actions autonomously, thereby transferring substantial operational risk and accountability to technology vendors and their AI-driven platforms. This transition is exemplified by Atera's deployment of their autonomous AI agent, Robin, which is positioned to handle a significant proportion of Tier 1 and complex Tier 2 IT tickets for managed service providers (MSPs). The company’s commercial strategy, including performance guarantees, signals an increased expectation that AI can assume core IT operational responsibilities that were traditionally reserved for human engineers. Atera has introduced a policy wherein Robin is guaranteed to autonomously close at least 50% of all Tier 1 and complex Tier 2 tickets within 90 days of onboarding, or fees are waived. According to Atera, this commitment is supported by a backend analysis of MSP tickets and live demonstrations using historical data. The company asserts that Robin’s mean time to repair is approximately 120 seconds, that onboarding is managed collaboratively, and that the rollout is more akin to hiring and training a human engineer than a standard software deployment. This approach is backed by patent filings and a business model integrating AI as the foundation rather than an add-on. The episode further examines the implications of mandatory AI bundling in Atera’s redefined RMM and PSA platform offering. The company has faced pushback from segments of the MSP community dissatisfied with bundled AI services and associated pricing changes, particularly from those wishing to maintain control over their technology stack. Atera responds by describing a re-conceptualization of their platform as inherently AI-driven, distinguishing between “platform AI” and the autonomous Robin agent, and clarifying that preexisting AI users would not incur additional costs. There is also discussion around the impact of automation on human roles and the need for new approaches to training and accountability, particularly for junior staff. For MSPs and IT service providers, these developments signal an increase in infrastructure dependency on vendor-managed AI agents, as well as new layers of contract risk linked to performance guarantees and platform integration. The operational reality described involves a significant reduction in required headcount, a shift in staff responsibilities from routine incident response to higher-order business and security tasks, and the necessity for designated internal management of AI tools. There remain unresolved concerns about skill degradation and the long-term risks of over-automation, including the narrower pathways through which junior personnel may acquire foundational experience. Sponsored by:  ScalePad https://scalepad.com/dave/ Nerdio https://nerdio.co/MSP-Radio Sign up for the SMB Online Conference: www.smbonlineconference.com &nbsp;💼 All Our SponsorsSupport the vendors who support the show:👉 https://businessof.tech/sponsors/&nbsp;🚀 Join Business of Tech PlusGet exclusive access to investigative reports, vendor analysis, leadership briefings, and more.👉 https://businessof.tech/plus&nbsp;🎧 Subscribe to the Business of TechWant the show on your favorite podcast app or prefer the written versions of each story?📲 https://www.businessof.tech/subscribe&nbsp;📰 Story Links &amp; SourcesLooking for the links from today’s stories?Every episode script — with full source links — is posted at:🌐 https://www.businessof.tech&nbsp;🎙 Want to Be a Guest?Pitch your story or appear on Business of Tech: Daily 10-Minute IT Services Insights:💬 https://www.podmatch.com/hostdetailpreview/businessoftech&nbsp;🔗 Follow Business of Tech&nbsp;LinkedIn: https://www.linkedin.com/company/28908079YouTube: https://youtube.com/mspradioBluesky: https://bsky.app/profile/businessof.techInstagram: https://www.instagram.com/mspradioTikTok: https://www.tiktok.com/@businessoftechFacebook: <a href="https://ww