Channel Brief: MSPs are all in on AI. The revenue still hasn’t followed
Managed service providers have raced into artificial intelligence. The experimentation phase is over, and customers have stopped asking which GenAI tool to buy. Now they want to know where AI belongs in their stack, who is accountable when something breaks, how it should be governed, and whether it’s actually moving the financial needle. That’s a tougher brief—and it’s where the real money should be. It just isn’t showing up yet.
New data underscores the gap between adoption and effective execution. GTIA reports that 97% of IT service providers are already using AI, but only 20% have the governance frameworks, formal policies, operating processes, and commercial strategies to run AI like a business. In short: nearly everyone is piloting; very few are industrializing.
Enterprise buyers are hitting the same wall. Domino Data Lab’s Fifth Annual Domino Enterprise AI Report finds 93% of AI leaders say the tech has improved production capabilities, yet 57% still don’t see returns exceeding costs—a number that hasn’t budged since 2025. The tools are in place. The payback isn’t.
The hard part starts now—and it’s an MSP-sized opportunity
Customers need help moving from “we have AI” to “we have accountable, safe, and ROI-positive AI.” This is squarely in MSP territory: stitching technology into workflows, building policy and controls, training people, securing data, and proving value. The services mix is shifting from platform resale to operational excellence.
High-value AI services MSPs can package now:
- Governance and risk frameworks: Policy development, model and tool inventories, RACI for AI decisions, vendor risk reviews, shadow AI controls, and change management.
- Secure data and identity foundations: Data minimization, tokenization and encryption, retrieval architectures, privileged identity for AI, approval workflows, prompt filtering, and data residency alignment.
- Workflow integration and automation: Connect AI to ITSM, RMM, PSA, knowledge bases, documentation, and runbooks; build evaluation harnesses; adopt multi-model strategies with task-specific model selection.
- FinOps for AI and measurement: Track unit economics (cost per ticket, per document, per lead), tag and report usage, set budget guardrails, and align KPIs to business outcomes.
- Workforce enablement: Role-specific playbooks, safe-use training, prompt hygiene, and “human-in-the-loop” operating patterns.
- Compliance and auditability: Log prompts and outputs, enforce retention and redaction, integrate DLP and content provenance, and prepare audit trails for regulators and customers.
- AI incident response: Red teaming and jailbreak testing, hallucination containment strategies, rollback plans, and post-incident reviews tailored to model behavior.
The MSPs that productize these services with clear SLAs, pricing, and outcome metrics will be best positioned to convert AI enthusiasm into durable revenue.
Channel moves and product updates
WatchGuard expands multi-model AI for MSP security
WatchGuard is broadening its frontier AI tooling via OpenAI’s Daybreak program and Anthropic’s Cyber Verification Program, giving its security teams access to multiple models for vulnerability research, product testing, red teaming, threat analysis, and remediation. The multi-model approach lets WatchGuard evaluate different systems and apply the right model to the right job instead of anchoring its security strategy to a single provider. For MSPs, that can translate into more resilient detections, faster response workflows, and better coverage across varied customer environments.
Absolute Security lands on the Pax8 Marketplace
Absolute Security has launched its cyber resilience platform on Pax8, opening a streamlined path for MSPs to deliver endpoint visibility, control, security, and remote recovery. The platform is designed to keep critical tools—RMM, EDR, VPN, identity services—operational and recoverable, reducing downtime and service disruptions. For partners, the Pax8 integration simplifies procurement, billing, and lifecycle management across multi-tenant deployments.
Keepit brings independent SaaS backup and recovery to Pax8
Keepit’s SaaS data protection platform is now available through Pax8, providing independent backup and recovery for Microsoft 365, Entra ID, Google Workspace, Salesforce, Okta, Jira, Confluence, and DocuSign. By storing backups in a separate, immutable cloud rather than alongside the production SaaS provider, Keepit helps customers harden against ransomware, accidental deletion, and provider-side outages—while improving compliance posture and time-to-recovery. MSPs gain a straightforward way to standardize backup across customers and apps.
Knowledge Grid debuts an AI-native data layer for cybersecurity
Knowledge Grid introduced the KG Cognitive Data Platform, a security data infrastructure layer that prepares raw telemetry for machine reasoning. The company argues that traditional SIEM and log platforms were built for human analysts, not AI models—often leading to false positives, missed detections, and overly confident but wrong conclusions. Built on its patented Temporal Data Grid and more than 15 years of research, the platform processes data at ingest to support downstream threat detection and agentic workflows. For MSP security operations, this could reduce noise, sharpen signal, and improve the quality of AI-driven decisions.
Point5 joins Sandler Partners to tackle complex connectivity
Point5 Managed Services has joined the Sandler Partners portfolio, bringing managed connectivity expertise to technology advisors, VARs, and MSPs handling complex, multi-site, or regulated-industry accounts. Point5 covers design, carrier sourcing, implementation, and post-sale support across data and voice optimization, colocation, MPLS migrations, and broader network transformation. The partnership targets projects that frequently stumble in delivery—especially global rollouts and deployments in healthcare, finance, and pharma—so partners can offload risk and keep timelines on track.
The bottom line
AI adoption isn’t the problem; monetization and operational discipline are. Customers already have models and copilots embedded across their stacks. What they don’t have is a governed, secure, and measurable AI operating model. MSPs that deliver that layer—policy, process, integration, security, and proof of value—will move from pilots to profit.
Have news to share or just want to connect? Reach us anytime at [email protected].
