The dominant structural shift outlined in the episode is the destabilization of the classic per-seat MSP bundle caused by the rise of agentic AI and token-based, metered automation platforms. Vendors such as Kaseya, Google, and OpenAI are embedding persistent AI agents within core business applications, moving beyond traditional licensing models to charges based on actions, tokens, and workflow usage. This introduces margin instability, as MSPs cannot reliably predict costs or maintain flat-rate contracts in an environment where AI consumption is dynamic and externalized.

The most consequential evidence presented is the quantification of AI-driven inefficiencies and costs in operational terms. According to a Gallup poll, cited by ZDNet, half of US employees are now using AI at work, but those users waste up to eight hours weekly managing AI-related tasks—amounting to approximately $1.25 million drag per year for a 100-person firm. This data underlines how the proliferation of automation does not equate directly to labor savings and can introduce significant, unanticipated costs that are difficult to contain under legacy MSP pricing models.

Supporting developments further highlight the governance gap and operational risk. Reports from PRWeb and Ruist find that 97% of MSPs intend to automate more in 2024, but only 4% are “highly mature.” Vendor announcements—as with Kaseya’s agentic IT management platform, Auvik’s Aurora AI agents, and Liongard’s data control enhancements—are paired with warnings from Information Week and The Register about the risk of overspending, audit failures, and accountability gaps tied to AI-driven automation. Most IT managers lack full control over AI agents, and as agents proliferate, the difficulty of tracking, governing, and assigning accountability rises.

For MSPs and IT service providers, these changes demand immediate attention to contract structure, governance, and pricing. Flat-rate, all-you-can-eat support models expose providers to untracked vendor consumption and hidden overages, making traditional agreements economically unstable. Practical safeguards require shifting toward consumption-based or outcome-based billing, enforcing explicit usage caps, audit controls, and vendor SLAs that clearly define liability and accountability. Failing to adapt risks absorbing uncontrolled automation costs and shouldering client disputes over AI-driven actions and expenses.

00:00 AI Overhead Crisis

04:48 Agent Control Gap

07:17 MSP Margin Squeeze

12:00 Why Do We Care?

Supported by:

Acronis
Zero Networks
Nerdio

Upcoming event:

The Pivotal Point of IT: Building Services for the AI-First Era
Date: May 13 at 1p.m. EDT
Register: https://go.acronis.com/davesobelaiera