The episode reveals a growing governance gap as the central structural shift in the IT services sector, driven by accelerated AI adoption and increasing automation. Companies such as OpenAI, Anthropic, Veeam, and Auvik are reframing their market positions around the operational risks and requirements introduced by AI agents, data automation, and new service delivery models. This evolution is underscored by the rising number of AI agents—projected by IDC to reach 2.3 billion by 2030—operating largely outside of current oversight and frequently with excessive or inappropriate permissions.

The principal development discussed is Veeam’s announcement of its Data AI Command Platform. According to Dave Sobel and Rich Freeman, this platform is intended to address data-centric failures beyond traditional ransomware or accidental deletion. Veeam’s platform is designed to handle issues such as AI-generated data hallucinations, inappropriate data exposure, and policy enforcement failures. The platform’s architecture builds on the acquisition of Security AI, combining data security posture management with backup, compliance, and governance capabilities, although, as of now, key remediation features are only available for Microsoft 365, with further expansion expected over the coming months.

Supporting developments include Auvik’s expansion of automated network management based on a large historical dataset and the simultaneous entrance of OpenAI and Anthropic into direct services for mid-market clients, backed by billions in private capital from entities such as Goldman Sachs and Blackstone. Both companies now embed applied AI engineers at client sites, bypassing traditional channel partners. Channel operator feedback, reflected in research by Techisle and discussions at vendor conferences, indicates a lack of MSP readiness and a slow response to developing governance and compliance services, despite evidence from end-user data pointing to significant unmet demand and risk exposure.

Operationally, MSPs face a growing liability trap where the speed and delegation of decisions to AI systems increase the potential for unnoticed errors or breaches. There is a disconnect between customer demand for governance, compliance, and data controls, and the preparedness of MSPs to deliver those services. This exposes providers to heightened contractual, operational, and reputational risk, particularly as vendors and large AI companies move directly into the mid-market service delivery space. Practical safeguards, clear accountability frameworks, and objective benchmarks for automation and governance effectiveness will be required to mitigate exposure and support safe, durable service offerings.

Supported by:
CometBackup
HaloPSA
Moovila