The Insurance Services Office (ISO) has introduced a new endorsement, CG 40 47 01 26 — Generative Artificial Intelligence Exclusion, that's beginning to appear on commercial general liability (CGL) policies. The endorsement removes coverage for certain losses arising from the use of generative AI, including tools that are increasingly embedded in routine construction activities.
As construction companies adopt AI-enabled solutions to support estimating, scheduling, safety, design coordination and project communications, this endorsement may create coverage gaps that aren't immediately obvious. These gaps can affect claim outcomes, contractual compliance, bid responsiveness and prequalification with owners and public entities.
This alert outlines what the exclusion does, why it matters for construction organizations, where exposures may arise and practical steps to consider.
What is the ISO Generative AI Exclusion?
The CG 40 47 01 26 endorsement excludes coverage under the CGL policy for bodily injury, property damage and personal or advertising injury arising out of, or attributable to, generative artificial intelligence.
Key features of the exclusion include:
- Broad wording that may apply even when AI is only one contributing factor to a loss
- Applicability to both Coverage A (bodily injury and property damage) and Coverage B (personal and advertising injury)
- Potential applicability whether AI is used directly by the insured or indirectly through vendors, consultants, or project systems
If a loss is alleged to be traceable, directly or indirectly, to generative AI output, the insurer may deny coverage under the CGL.
Why does the ISO Generative AI Exclusion matter for construction companies?
Everyday uses of AI may trigger unintended exposure
Many construction organizations already rely on AI-enabled tools, including:
- AI-assisted estimating and quantity takeoffs
- Schedule optimization and sequencing tools
- Safety analytics and predictive hazard identification
- AI-generated specifications, narratives, or project communications
- Design support, modeling, or clash detection tools
These technologies are often viewed as decision-support tools rather than risk-driving systems. Under the exclusion, however, the presence of generative AI anywhere in the causal chain may be sufficient to challenge coverage.
Illustrative example: An AI-assisted shop drawing or calculation contains an error that contributes to a structural failure or injury. Even if project personnel reviewed and approved the drawing, the insurer may argue that the loss arises out of generative AI and is therefore excluded.
Heightened implications for public entity and complex projects
The exclusion may be particularly consequential for:
- Public entity projects
- Large or multi-phase construction programs
- Contracts with strict insurance and indemnification requirements
Many owners and public agencies continue to assume that a standard ISO CGL policy provides broad protection for bodily injury and property damage. The introduction of this endorsement can create a disconnect between contractual insurance requirements and actual policy terms, potentially leaving contractors exposed despite apparent compliance.
Contractual, bid and prequalification considerations
Most construction contracts and insurance specifications:
- Don't address AI usage.
- Don't contemplate AI-related exclusions.
- Rely on long-standing assumptions about CGL coverage scope.
As a result, organizations may unknowingly:
- Accept uninsured risk.
- Breach insurance-related contract obligations.
- Create downstream gaps in subcontractor risk transfer.
These issues often surface only after a loss occurs or during claim review, rather than at policy issuance.
Practical steps to consider
While the insurance market's response to AI risk continues to evolve, construction companies can take several practical steps now.
Understand how AI is used across your organization
- Identify where AI-enabled tools are in use (estimating, scheduling, safety, design support, communications).
- Distinguish between decision support, delegated decision-making and technical substitution.
- Understand whether AI tools are internally developed, vendor-provided, or embedded within third-party platforms.
Establish internal AI governance and controls
- Define approved tools and acceptable use cases.
- Set expectations for human review, validation and documentation.
- Establish guardrails for external communications, contract comments and submittals.
Clear governance can reduce both operational risk and uncertainty in the event of a claim.
Review insurance programs and endorsements
- Confirm whether CG 40 47 01 26 or similar language has been added to your CGL program.
- Assess alignment between AI-related exclusions and actual operational practices.
- Explore whether endorsements, manuscript carve‑backs, or complementary coverages may be appropriate.
Align contracts with current insurance realities
- Review owner contracts for assumptions about CGL coverage.
- Consider whether subcontract language should address AI-related exclusions or alternative risk transfer mechanisms.
- Evaluate whether additional disclosures or risk allocation provisions are warranted.
Looking ahead
The ISO Generative AI Exclusion represents an important shift in how insurers are responding to emerging technology risk. For construction companies, the issue isn't whether AI will be used, but whether its use is recognized, managed and aligned with insurance and contractual obligations.
Addressed proactively, this issue can help reduce uninsured exposure, avoid surprises at claim time, preserve bid competitiveness and support defensible risk transfer strategies.
If you have questions about how this exclusion may apply to your operations, projects, or insurance program, we recommend a focused discussion to review:
- Your specific AI use cases
- Current policy language and endorsements
- Alignment between contracts and insurance coverage
As AI continues to shape construction operations, aligning its use with insurance and contractual protections is increasingly important. Handled thoughtfully, this issue can be addressed without disrupting operations while strengthening overall risk governance. To discuss your approach, reach out to our team for a focused review.
