MSPs will invest in more AI security forecasting

We have already seen the adoption of anomaly detection for cyber threats and we will see more of this in other products. In my opinion, the use of AI for security forecasting will become its own field of study and its own set of products, moving from a feature within a product, to its own product.

End users will get pickier when it comes to AI solutions

Customers will start getting pickier with AI demands in 2025. When we first saw LLMs emerge two years ago, it wasn’t that big of a deal if a prompt didn’t return anything useful – it was a new technology, and people were just happy to play around with it. Now, customers are going to start demanding accuracy and start pushing for ethical oversight.

Across the industry, we’re already starting to see informed customers referring to generative AI as a security and ethics risk. To achieve reliable results, AI has to have enough information about users – but when you have this much information, security questions arise. Customers will demand transparency from MSPs, with clear explanations of AI usage, before they accept its use.

As customers push for this, vendors will prioritise AI tools with strict data privacy and security protocols to meet current and upcoming legislation.

AI talent will be in shorter supply

MSPs are already struggling with a talent shortage, and it’s not going to get easier if they need to hire AI specialists. Talent with AI and LLM skills are in high demand and have high earning expectations. MSPs want to build their own agents and customer interfaces, or they may want to adopt AI built by others, and that will require AI understanding, prompt engineering skills, and large language model knowledge. Rather than hiring externally, MSPs need to think about their own employee base, and upskill and train their existing staff. They already have a technically adept workforce, so they have a head start.

We will also see MSPs leaning heavily on vendors to make up for a lack of AI expertise, demanding prebuilt, no code tools which create a workaround for talent shortages.

AI incident planning will become a staple

Much like we have security tabletop exercises now, for ransomware we can expect to see incident response plans to mitigate AI failures. MSPs will first conduct exercises themselves and then take this concept out to customers.

Most MSPs are aware of the risks AI poses, but there are still some that use it without fully appreciating its issues and use it haphazardly. This will change very quickly in 2025 with new regulations to make sure IT organisations use AI safely. Part of meeting this regulation will be making sure there are incident response plans that treat the issue as seriously as any other breach.

MSPs Will Build and Adopt Task-Oriented AI Agents

One of the most transformative opportunities for MSPs in the AI space lies in the development and adoption of AI agents. These are task-oriented components integrated into generative AI solutions that enable automation of specific actions in meaningful and impactful ways. Unlike broader AI systems, these agents are finely tuned for distinct tasks—ranging from proactive monitoring and ticket resolution to predictive analytics and compliance checks.

MSPs looking to enhance their internal operations or deliver more intelligent solutions to their customers will increasingly rely on these AI agents. By building AI agents tailored to their unique workflows and business models, MSPs can address challenges such as resource optimization, reducing manual workload, and improving service quality.

For vendors, this shift represents a call to action. Much like N-able’s Ecoverse and its open platform approach, which opens APIs for data access and action enablement, vendors will need to adopt strategies that facilitate seamless integration of AI agents with their products. Supporting task-oriented agents will mean providing not only robust APIs but also frameworks for secure, efficient interaction between AI and the underlying systems.

This evolution underscores a larger trend: AI isn’t just about automation—it’s about smart, actionable automation that aligns with business goals. MSPs and vendors who embrace AI agents as part of their generative AI ecosystems will be well-positioned to lead in an increasingly competitive landscape.