Episode #914 of the MSPi PrimeCast
AI is moving quickly into small and midsize businesses, but the controls surrounding it aren’t always keeping pace. Saurabh Sandhir, founder of Kipling Secure, believes SMBs may actually adopt AI faster than larger enterprises because they have fewer internal processes standing in the way.
For MSPs, that creates a new responsibility. As the technology and security partners many SMBs already rely on, MSPs can help customers use AI more thoughtfully while addressing the risks that come with rapid adoption.
Here are eight takeaways from Sandhir’s perspective that MSPs can put into practice.
1. Bring Shadow AI Into the Open
Businesses may think employees are using only a few familiar AI platforms. Sandhir says the reality can be much broader. AI capabilities are increasingly embedded within applications, creating a collection of unknown or unvetted AI tools with different models and data-sharing policies.
MSP Action: Help customers identify the AI and AI-enabled tools being used across their environments so they can make informed decisions about what should be permitted.
2. Protect Sensitive and Proprietary Data
Employees don’t have to act maliciously to create AI-related data risk. Sandhir points out that employees trying to become more productive can share PII, PHI, financial information, customer data, or proprietary business information with AI systems.
MSP Action: Help customers establish clear policies and controls around the types of information employees can share with AI platforms.
3. Prepare for AI-Specific Security Threats
AI can make businesses more productive, but Sandhir notes that attackers can benefit from the technology as well. He describes emerging AI-oriented attacks and vulnerabilities that differ from the threats businesses are accustomed to managing.
MSP Action: Begin incorporating AI-specific risks into customer security discussions and assessments as AI becomes more deeply integrated into business workflows.
4. Address Toxic and Misleading AI Content
Not every AI risk involves a cyberattack. Sandhir shares an example of a voice-to-email tool that worked correctly most of the time but occasionally inserted gibberish into business emails.
Errors like these can create reputational problems even when the technology is working as intended most of the time.
MSP Action: Encourage appropriate human oversight for AI-generated content, particularly when that content reaches customers, partners, or other external audiences.
5. Put Guardrails Around AI Without Slowing Innovation
Sandhir doesn’t advocate blocking AI. He believes in the positive impact the technology can have. His focus is on putting the right guardrails around its use so businesses can gain the benefits without allowing adoption to become uncontrolled.
He describes the goal as smarter AI adoption rather than ad hoc AI adoption.
MSP Action: Position AI governance as an enabler of safer AI use rather than simply another set of restrictions.
6. Connect AI to Real Business Outcomes
Sandhir believes businesses should ask more than which AI tools they are using. They should also determine which workflows they want those tools to influence and what outcomes they hope to improve.
In his view, AI should be a means to an end rather than an end itself.
MSP Action: Start customer AI conversations with the workflow or business problem first, then determine where AI can appropriately support that objective.
7. Keep AI Governance Simple and Practical
Working with MSPs reinforced for Sandhir how important simplicity is. He describes MSPs as a “forcing function for simplicity” because the solutions they deploy must work efficiently across many SMB customers.
That lesson matters as AI governance becomes another responsibility MSPs may need to manage.
MSP Action: Evaluate AI governance solutions based on how well they fit existing MSP operations, customer environments, and workflows—not simply on the number of features they offer.
8. Turn AI Governance Into a Recurring MSP Service
Sandhir sees an opportunity for MSPs to go beyond simply absorbing the risks created by customer AI adoption. He believes AI governance and AI detection and response can become services MSPs provide to their customers.
That can allow MSPs to help protect and control customer AI use while potentially creating new MRR and ARR.
MSP Action: Consider where AI governance and AI detection and response could fit within your managed services portfolio as a defined recurring offering.
What This Means for MSPs
Sandhir describes MSPs as the “operating systems” for SMBs and the real CIOs, CISOs, and help desks for businesses that often lack those resources internally. That puts MSPs in a natural position to help customers navigate AI adoption.
The opportunity isn’t about stopping AI. It’s about helping customers understand what they’re using, protect sensitive information, recognize new threats, and put practical controls around the technology.
As AI becomes more deeply embedded in SMB operations, MSPs that can bring greater visibility, security, and governance to its use can strengthen their role as trusted technology partners—and potentially build new recurring services in the process.
Catch the full conversation on MSPi PrimeCast Episode #914 and connect with Saurabh at https://www.linkedin.com/in/ssandhir/