AI may seem to replace knowledge work – and yet, at that very moment, it makes human expertise the industry’s most valuable asset.
Every wave of innovation in the history of IT has changed something: products, platforms, delivery models. None has truly touched the underlying business model. AI is different. Above all, the evolution from generative AI to agent-based AI – that is, the leap from AI that advises to AI that acts – opens up entirely new operational applications for companies and, with them, raises questions about accountability, governance, and trust.
The first wave of experimentation, proof-of-concepts, and internal pilot projects is over. What is beginning now is the implementation phase. And it raises different questions: not whether AI works, but whether it is economically viable. NetApp CEO George Kurian put it this way during a conference call with Wall Street analysts: “AI is no longer a future aspiration. It’s a business imperative.” Customers are realizing that data accessibility, governance, and preparation must come before AI. The infrastructure, not the model, is the bottleneck.
Antonio Neri, CEO of HPE, describes this same shift from an operational perspective. Today, customers are no longer primarily asking about AI models, but rather about the economic implications, such as infrastructure costs, governance efforts, and operational efficiency. Networking, storage, and hybrid cloud are just as strategic as the AI layers themselves. For MSPs and MSSPs, agent-based AI further intensifies these questions: Who controls what an AI agent decides? Who bears the responsibility? Who orchestrates it? As one IT system house manager puts it, these are “the major topics of discussion with our customers.” And they are questions to which no software vendor provides a complete answer.
The channel is shifting from a technology marketplace to a talent marketplace
This is where the real change for the IT channel lies, and it is more structural than anything that came before.
Every previous wave of innovation, such as client-server, the cloud, or SaaS, brought new products into the channel. The business models behind them remained recognizable: reselling licenses, implementing projects, and signing maintenance contracts. AI breaks this pattern because the barriers to its adoption are no longer technological. Those who implement AI in their companies do not fail because of computing power, but because of inaccessible data, undocumented processes, governance gaps, and a lack of security policies. The complexity lies in the organization, not in the technology.
This results in a shift in the purchasing decision and an opportunity for channel partners to move up the value chain. Customers no longer ask, “Who has the most comprehensive vendor portfolio?” They ask, “Who has the people who know what to do with it?” AI architects, security specialists, data engineers, and transformation consultants are the scarce resources. Today, the fiercest competitors for channel partners are not other resellers and solution providers, but hyperscalers, vendors, consulting firms, and enterprise IT departments.
This is the market shift: from technology to talent. The most successful partners in the coming years will not be those with the largest vendor portfolios, but those who have built world-class teams with AI expertise: teams that can make any product a success and help customers safely adopt AI, effectively integrate it, and demonstrate concrete business results.
The Question That Cannot Be Delegated
Stefan Fritz precisely articulates the operational implication. For every business owner introducing AI into their company, there is a critical question that cannot be delegated: Can someone who hasn’t done the work themselves determine within a reasonable amount of time whether the result is correct?
In the service and knowledge-based sectors, the honest answer is almost always no. And that points the way forward. The key lever lies not in purchasing models, but in breaking down one’s own work into verifiable components. This is process work. Only this time, it’s not between people, but between people and machines. Those who can do this make AI operational for customers. Those who cannot make promises without results.
That is precisely the competitive advantage that AI does not provide: the judgment, experience, and architectural work that make AI meaningful in the first place. After all, as AI becomes more autonomous and increasingly integrated into operational workflows, success depends less on access to the most advanced models than on the ability to effectively manage, secure, and deploy them.
In the implementation phase, the winners won’t be those who get the latest model the fastest, but those who know what to do with it.
The question facing channel partners
The AI era is transforming the role of IT partners and further driving the shift from being purely technology providers to strategic knowledge partners. The challenge lies in adapting their own business models and developing new service offerings in which AI expertise can have its full impact.