Here’s a number that should make everyone uncomfortable: it takes the average organization 241 days to identify and contain a breach. Meanwhile, the time-to-exploit for a new vulnerability is down to about a day.
That gap is the reason we built the Global AI Security Readiness Report. We surveyed 800 MSPs and 800 in-house security teams across nine countries to find out teams are preparing for the future of AI-driven threats.
The same attackers, two very different levels of readiness
MSPs and in-house teams are facing identical adversaries. Same techniques, same pressure to protect the business. But when you look at how ready each group is to fight back with AI, the picture diverges fast.
- 71% of MSPs say they’re very confident in their ability to detect and respond to incidents. Only 43% of in-house teams say the same.
- MSPs are using AI across security operations at a 60% average rate. In-house teams are sitting at 44%.
- In-house teams reported or suspected a security incident at nearly 3x the rate seen among MSP clients: 60% versus 22% on average.
None of this means in-house teams are behind because they’re not trying. Budgets are moving: 78% of in-house teams and 89% of MSPs plan to spend more on security tools in 2026 than they did last year. The issue isn’t willingness. It’s pace. Attackers are moving at machine speed, and defenders, even well-funded ones, are still moving cautiously.
What everyone agrees is the biggest threat
Ask an MSP and an in-house team what they’re most worried about defending against this year, and you’ll get almost the same answer: AI-enabled phishing and social engineering (39-40%), followed closely by identity-based attacks like credential theft and MFA fatigue.
That alignment matters. It means the conversation between MSPs and their customers doesn’t have to start with “here’s what you’re missing.” It can start with “we already agree on the threat, let’s talk about the gap in how we’re each equipped to stop it.”
And that gap is real. 59% of in-house teams want to involve their MSP when adopting new cybersecurity capabilities. For agentic AI specifically, nearly two-thirds would rather bring it in through an MSP or a hybrid model than build it alone in-house.
Why MSPs are pulling ahead
Part of it comes down to volume and variety of exposure. MSPs see a much wider range of incidents across their client base, which builds a deeper bench of pattern recognition than any single in-house team can develop on its own. Ransomware is the top incident type MSPs report handling, while in-house teams face business email compromise most often. That difference alone says a lot about the breadth of what MSPs are up against day to day.
MSPs are also managing more tools (8 on average, versus 5 for in-house teams), but that’s not automatically a good thing. 78% of organizations say tool sprawl is actively hindering their ability to mitigate threats. More tools without a unified platform just means more noise and more places for something to slip through. The advantage isn’t the tool count, it’s that MSPs are more likely to be consolidating that sprawl into a platform that sees the full attack path instead of isolated alerts.
The takeaway I’d want every reader to leave with
AI readiness is not a future aspiration anymore. That’s not a scare tactic, it’s what the data shows, and it’s echoed by the guidance coming out of international government agencies.
In June, the US issued an executive order pushing public-private collaboration on AI-enabled defense. Days later, the Five Eyes cyber agencies warned that frontier AI will reshape both offensive and defensive capability on a timeline measured in months, not years.
We don’t have the luxury of a long runway to decide whether AI belongs in our security operations. We’re running on borrowed time to figure out how we will respond as AI changes the threat landscape forever.
Threat actors are already operating at machine speed, aided by AI tooling and porous defenses. The organizations pulling ahead right now, MSP or in-house, are the ones treating vulnerability exploits, phishing defense, identity protection, and data protection as one connected strategy in the race against advanced chains of attacks.
Want the full data set? Download the Global AI Security Readiness Report to see the regional breakdowns, the full incident and tooling data, and where the biggest AI readiness gaps are opening up in 2026.