EDGE 2025: AI demands MSP transformation – but into what? 

Nobody is better positioned to solve the problems of operationalising artificial intelligence (AI) for small to medium-sized businesses (SMB) than managed service providers, but the reality may be that they need help and clarity to offer real transformation. 

During the second day of EDGE 2025, Pax8 chief product officer Libby McIlhany said in her presentation Capitalising on AI to Grow into the Era of Managed Intelligence that today’s managed service providers are going to be providing the context for the AI that they’re going to be delivering to SMBs. 

“You know your SMBs’ businesses as well as they do,” she said. “You’re going to be having those deeper conversations with them and you’re going to be providing the data and the information that AI agents need to be trained on.” 

McIlhany put to the MSPs that they were going to turn those abstract models into actionable business online systems. 

“You’re already going to lean into the two languages that you speak today — … the language of business and the language of technology,” she said. 

The industry is now at an inflection point between managed service providers (MSP) and managed intelligence providers (MIP), she noted. The MIP is different from traditional MSPs because they will have a higher labour force of humans and of human understanding. 

“You’re going to be overseeing agents or addressing clients,” she said. “You’re the one who’s going to help the clients understand how to embed AI into their workforce so that they can get not only more efficient, but also more capable. 

“You’re going to be delivering outcomes for their business, and you’re going to act as a strategic intelligence leader, because they don’t want to do it.” 

While Mcllhany painted a vision of the evolved channel partner, TSP Chief Strategy Officer James Davis said MSPs still faced gaps to getting there. 

The path to evolved MSP is not paved with gold 

Before an MSP even considers coming down this route, Davis said MSPs who have come from an end-user support and infrastructure management background, need to consider what their transformation looks like. Otherwise going straight into becoming an MIP was going to be a big leap. 

This is something that Davis has often spoken about. Previously he told ARN smaller MSPs need a “mindset shift” and start seeing the industry as a place for transformation and innovation, not just maintenance and support. 

The evolution of MSPs “started in as break-fix shops” then moved to infrastructure management, but now basic infrastructure and desktops are self-servicing, Davis said at EDGE Day 2 in an interview. 

“For decades, MSPs have specifically avoided the application layer in their clients’ businesses,” he said. “They don’t know that clients need advisory first to lead them, and they don’t actually know what they’re trying to solve. 

“They can’t see the wood from the trees — it’s not technology.” 

Davis noted that AI is a conversation that requires talking about productivity and efficiency. 

“However way we solve that for our clients doesn’t actually matter,” he said. “But they’re struggling to scale, struggling to recruit and retain talent, and they’re struggling to create profitability.” 

“We don’t — and can’t — come to the AI conversation with an off-the-shelf product for customers, trying to solve a problem that we don’t know.” 

“The way the AI is being positioned to MSPs is the wrong way around,” said Davis. While he doesn’t want to be pedantic, even the word ‘provider’ for Davis begs the question: what are they providing? 

“Not all partners are building solutions — they’re procuring — so automatically, we’re getting off on the wrong foot because we’re not understanding the business models,” he said 

Partners know they need to transform, but they don’t know what they should transform to. A lot of them are also in survival mode and are desperate because they are trying to keep their business profitable. 

“They don’t know what the steps of transformation are actually like,” he said. “They don’t have the capabilities. 

“They’re not doing anything like this themselves — and they haven’t even got cyber security.” 

The shift being discussed, along with the language used by vendors, noted Davis, misses the mark. 

“Vendors, even with good intentions, are focused on selling licenses rather than understanding the broader implications,” he said. “This misunderstanding is evident in how they talk about agentic AI. 

“Even if they’ve got the right intentions, they don’t actually understand the implications of all this.” 

True partnership takes time 

An average MSP is horizontal, explained Davis, and they can’t go deep in their verticals and their clients, because they have a variety of verticals that they have already to work with. 

“There are really good solutions across machine learning, edge devices, [and] specialised, intelligent AI that makes decisions based on a certain subset of data that’s trained very well and very specifically,” he said. “They exist, but that’s not what we — as an industry — are about. 

“That’s why the partners need to transform, because if you’re just following a commodity, their customers don’t need someone to come in and implement a commodity-grade solution.” 

Their customers need a partner that can understand their business practically, which requires potential solutions to help them implement AI and govern it, noted Davis. 

“It’s not about helping a partner understand whether they should be elevating to advisory status or getting very specialised in technology,” he said. “Even the term MSP to equal all partners — that’s a ubiquitous term and it loses all meaning. 

Partners might get a bit of buzz in the first place. However, as soon as they walk away after the implementation, the concept dies. 

“They don’t do anything — surprise, surprise — they don’t understand it and they don’t believe in it,” Davis said. 

Channel Guru founder Brad Clarke said partners need a community of experts that they can lean on to build competence that gives direct access to experts. 

“You can teach an old dog new tricks… if the dog wants to learn,” he said. “You’ve got the old dog who doesn’t want to learn, or doesn’t care to learn, is burnt out and over it — they’re your exits. 

“Then you’ve got those who want to transform — and there’s plenty of great communities out there who are focused on very much that.” 

However, they’ve got to want to transform, Clarke pointed out. 

“For example, it can take up to six months for someone like James [Davis] to help them understand what a modern version of that part of the business will look like — and they’re going to take another three or four months to implement,” he added.