The AI Gold Rush is here. Are MSPs prepared?

MSPs are widely deploying AI-based services and tools, but more training and a renewed focus on proven, reliable, and straightforward applications is needed.

Patrick O’Donnell, SVP, Worldwide MSP Sales, Barracuda

Artificial intelligence (AI) has rapidly gone from a buzzword to reality across many industries and applications, and managed service providers (MSPs) have had to quickly pivot to address emerging demand and the security challenges that AI presents.

According to Barracuda’s seventh annual MSP Day report, MSPs have rapidly adopted AI, but most feel like they need more education while simultaneously feeling pressured to meet the demand for these solutions.

Enhancing Cybersecurity and Efficiency: The Power of AI in Managed Services 

The annual Barracuda report found that AI-based tools, products, and services can reduce costs and resource demand for MSPs and their clients through task automation and scaling. AI and machine learning can also improve cybersecurity by helping MSPs provide 24/7 monitoring and notifications by quickly recognizing application or network behavior anomalies that would be difficult for staff to see independently. 

Barracuda’s Managed XDR offering uses AI and machine learning to monitor, detect, and respond to cyber threats at any time of day or night. In Microsoft 365 environments, for example, Barracuda uses 90 days of data to analyze logins and detect suspicious activity. Given its advanced features, Barracuda XDR can achieve a 90 percent true positive rate for machine learning (ML) rules due to its high confidence and continuous AI learning capabilities. From January through July 2024, Barracuda XDR analyzed nearly 24 million Microsoft 365 events and neutralized 387 high-risk incidents.

Furthermore, Barracuda Intrusion Detection Systems (IDS) are crucial for detecting security breaches, particularly through the “High Destination Bytes Transferred” alert. This metric identifies unusually large data transfers within an organization, aiding in the detection of possible threats by monitoring and analyzing network traffic patterns. Alerts are triggered when data transfers exceed normal baselines, potentially indicating malicious activities like data exfiltration. Between January and July 2024, Barracuda IDS analyzed over 250 million events and neutralized 15 high-risk incidents associated with exfiltration.

In addition to that, Barracuda Email Protection offers a multi-layered approach combining a secure email gateway, AI-powered fraud protection, advanced security awareness training, and automated incident response. With 91% of cyberattacks starting with an email, it is critical for MSPs to be able to proactively monitor their customers’ email security postures and defend against spear phishing, business email compromise (BEC), account takeover attacks, domain impersonations, and more. 

Finding the Right Balance: Matching Skill with Customer Demands

According to Barracuda’s research, ninety-five percent of MSPs responding to the survey currently use AI (58 percent) or plan to onboard AI services in the next year (37 percent). Common AI-powered services include enhanced cybersecurity, intelligent automation and efficiency, predictive analytics and proactive solutions, and personalized training and education.

Yet, many MSPs still feel ill-equipped to help their clients with these AI-powered solutions. According to the report, nearly 9 out of 10 MSPs (87 percent) said they need significant or notable improvements in their knowledge and application of AI, and 30 percent reported that enhancing their in-depth understanding of AI is a top challenge. This is important because 77 percent reported feeling pressured to offer clients AI insights and tools. 

Beyond the Hype: AI Implementation Success

According to the Barracuda report, researchers found additional drivers of AI in managed services include better cybersecurity, improved efficiency and resource allocation, and integration of predictive maintenance. 

MSPs that have incorporated AI into their services report using it in predictive analytics and proactive solutions (57 percent of those surveyed), followed by intelligent customer support (55 percent) and enhanced cybersecurity (54 percent). Significantly, almost every respondent not currently offering these in their services plans to add them in 2024. 

The need for education and caution remains, however. MSPs and their clients can benefit from an organized approach to AI utilization. AI is best deployed to augment human activity, not replace it—the industry has settled on AI as a co-pilot for many applications. In other words, final analysis and decision-making should be handled by experts, guided by well-trained AI algorithms sifting through clean, reliable data.

When improperly deployed, AI can quickly prove to be costly and useless. Large language models (LLMs) trained on massive amounts of unfiltered data can hallucinate, and some tests have shown that if the models are fed AI-generated data, they can suffer from what is known as model collapse. The AI results then become unreliable and, in some cases, nonsensical. 

A report from AI search specialist Lucidworks illustrates some of the caveats of AI deployments. This global survey found that AI adoption is beginning to slow down, with just 63 percent of organizations planning to increase AI spending (compared to 93 percent in 2023). This is a result of more diligent prioritization and thoughtful planning. Furthermore, the study also found that just 25 percent of planned AI projects are fully implemented, and 42 percent have not seen any significant benefits from generative AI.

The Human-AI Partnership: The Future of Managed Services

AI is here to stay, and for MSPs struggling with client demands, a key strategy will be to invest in training and focus their efforts on proven, reliable, and straightforward applications that can augment and automate specific tasks. They should also help clients apply that same focus to their own deployments and caution them to clamp down on rogue deployments of unauthorized generative AI tools that could pose security risks. To make the most of the technology, MSPs also need to be practical in their implementation strategies, while keeping security top of mind. 

Patrick O’Donnell is SVP of Worldwide MSP Sales for Barracuda, where he is responsible for accelerating Barracuda’s MSP sales, driving the go-to-market strategy, execution, and programs for the sale of security, and data protection, and Managing XDR offerings through the channel. Connect with Patrick on LinkedIn.