SPECTRA launched an AI risk framework on September 10, 2026, designed to give managed service providers, small- and mid-market businesses, and insurers a common standard for managing risks tied to AI deployment. The move comes as insurers grow more cautious about which AI-related exposures they will cover, and as businesses adopt agentic AI faster than policy language can keep up.
The framework includes 20 AI deployment controls across eight domains. SPECTRA developed the controls with input from cybersecurity advisers, MSPs with established AI practices, and cyber re/insurers and brokers. Several MSPs are now working toward certification in a pilot phase.
Where AI coverage gets murky
Insurers are still sorting out how standard policies respond to AI-driven failures. A recent report from At-Bay Insurance flagged risk implications from agentic AI across multiple lines. General liability policies typically cover property damage and bodily injury, but often remain silent on AI-related claims. Professional liability policies show the same gap when failures stem from AI outputs.
Cyber insurance presents additional questions. At-Bay said policies should be reviewed for business interruption tied to AI service availability. Privacy breaches where an AI solution acts as the vector raise similar concerns. Copyright and intellectual property infringement from AI-generated content also sits in a gray zone - coverage may be unclear.
Responsibility for AI-related risks between MSPs and their customers remains another open question. The framework aims to clarify who manages what.
Twenty controls across eight domains
The framework identifies the specific controls businesses are asking their MSPs to handle. SPECTRA said the structure gives MSPs, businesses, and insurers a shared approach to mitigating AI deployment risks. The company operates at the intersection of managed cybersecurity and insurance, and has watched MSPs accelerate AI deployment while insurers become more selective about coverage.
SPECTRA is working with carriers to define security control requirements and the data needed to demonstrate those controls. The company is also collaborating with insurers on affirmative AI coverage - what carriers will cover after inspecting the controls.
Insurers want deployment-level data
Gage Caligaris, founder and CEO of commercial insurance platform Ledgebrook, said insurers need evidence and data about AI controls at the point of deployment. Ledgebrook insures tens of thousands of businesses nationwide.
"Greater visibility into how businesses deploy and govern AI allows insurers to more precisely assess and price the exposure," Caligaris said.
SPECTRA founder and CEO Edouard von Herberstein said the framework is intended to provide MSPs with a structure to help small businesses deploy AI while addressing guardrails, security, and financial resilience.
Why this matters for insurance professionals
Brokers and underwriters face growing pressure to clarify coverage for AI-related exposures. A framework with defined, certifiable controls gives carriers something concrete to inspect and price against - moving the conversation from theoretical risk to measurable deployment practices. For professionals working at the intersection of AI for Insurance and AI for Cybersecurity Analysts, the framework signals where underwriting standards are heading. SPECTRA's pilot with MSPs will produce the kind of deployment-level data insurers have been asking for, and early results will likely shape how policies address AI risk in the small and mid-market segment.
Your membership also unlocks:
