The Insurance Industry Is About to Discover AI Governance in Buildings — And It Won't Be Pretty

The Insurance Industry Must Address AI Governance in Building Operations

By James C. Waddell, President, Cognitive Corp

---

The Growing Challenge

As artificial intelligence (AI) continues to reshape building operations, the insurance industry faces a pressing challenge. The deployment of autonomous agents, such as those utilized for energy management and safety functions, is accelerating without sufficient governance frameworks in place. This oversight gap not only complicates risk assessment for insurers but also amplifies liability concerns that must be addressed urgently.

Recent analyses reveal that commercial property insurers are grappling with annual premium increases ranging from 5% to 15%, primarily driven by climate risks and catastrophic losses. With no established governance protocols for autonomous systems executing pivotal operational tasks, significant liability risks arise. Insurers must contend with scenarios involving safety violations, code breaches, or tenant injuries, leading to complex liabilities that their claims departments find increasingly challenging to navigate.

Insurers are now asking: "Can you provide insight into the decisions made by the AI system?" Regrettably, the absence of clear governance frameworks means these inquiries often go unanswered—a situation that is unsustainable as regulatory scrutiny intensifies.

Alarmingly, buildings employing ungoverned autonomous systems are becoming uninsurable by traditional standards, prompting a need for insurance carriers to reassess their underwriting practices. Those who proactively formulate comprehensive governance measures will position themselves as leaders in an evolving market landscape.

---

Understanding the Liability Gaps

The ramifications of operating autonomous agents in buildings are stark. For example, envision a facility manager using an energy optimization agent designed to reduce energy costs. On an unexpectedly cold day, this agent could reduce heat output by 18%, neglecting to monitor consequences in critical areas like server rooms. If the temperature drops dramatically, it could inflict millions in damage due to thermal shock.

When an insurer is notified of the damage, they will probe into operational accountability by asking:

• "Who authorized that decision?"

• "What parameters were established for the agent’s operations?"

• "Are there records explaining the decision that was made?"

The reality may reveal an unsettling truth: no documentation exists to clarify these autonomous decisions, marking a critical accountability void. Traditionally, human operators provide rationale and authorization for decisions; however, autonomous agents operate outside such frameworks:

• No human oversight is recorded.

• No reasoning for decisions is provided.

• Lack of established accountability.

• No documented evidence of decision validity.

From an insurer’s perspective, this creates an immeasurable risk leading to potentially significant financial liabilities. As this pattern proliferates across numerous properties employing hundreds of autonomous agents, the insurance industry will be compelled to seek clarity on these decision-making voids and emerge with innovative risk management practices.

---

Drawing Parallels with Cybersecurity

The intersection of evolving technology and insurance is not unprecedented. A decade and a half ago, cybersecurity risks were often largely ignored in underwriting assessments. Organizations like hospitals and banks were seldom asked about their cybersecurity preparedness.

This reality shifted dramatically following notable data breaches, including incidents where hospitals suffered crippling ransomware attacks. This catalyzed an awakening within the insurance realm, as underwriters recognized the necessity for visibility into cybersecurity measures to aptly evaluate risk. Cyber insurance evolved into a specialized category, with frameworks established to require and evaluate robust security protocols.

A similar momentum is now emerging in the realm of AI governance for buildings. As autonomous systems proliferate without sufficient oversight, insurers are beginning to confront the same blind spots faced in the early days of cyber risk. Achieving clarity and governance is essential for accurately gauging technological risk exposure. The first insurer to develop a comprehensive AI governance risk assessment framework is poised to set new market standards.

---

Outlining AI Governance in Buildings

To implement effective AI governance within building operations, three essential components must be established: explainable decisions, human oversight checkpoints, and auditable decision trails. We refer to this comprehensive framework as the Building Constitution.

Applying this to our earlier example of the energy optimization agent:

1. Explainable Decisions (XAI): Agents must operate under defined, verifiable rules:

• Rule 1: "Ensure heating levels stay above 62°F in critical infrastructure, like server rooms."

• Rule 2: "Maintain temperature above 55°F in occupied spaces when the external temperature is below 32°F."

• Rule 3: "Limit heating reduction to a maximum of 12% in any given hour."

Such parameters facilitate audits by third parties verifying compliance with safety standards.

2. Human Oversight Checkpoints (HITL): For any critical decision impacting infrastructure, approval from a human operator is required. While not every action necessitates this oversight, those with implications for safety must undergo rigorous review. The example of the energy optimization agent necessitating approval from the facility manager exemplifies this.

3. Auditable Decision Trails: Each choice made by an autonomous agent must be logged systematically, offering an accessible audit trail for insurers during claims processing. If an incident arises, insurers should be able to reference every pertinent decision made regarding critical infrastructure, ensuring accountability through documented decision paths.

The adoption of these governance measures is poised to transform insurance operations, with the associated implementation costs for integrating such frameworks into building systems being manageable within four to eight weeks.

---

Introducing the Building AI Governance Index (BAGI)

Cognitive Corp is excited to announce the Building AI Governance Index (BAGI), a metric evaluating governance maturity across seven dimensions aligned with the Building Constitution: Safety, Transparency, Fairness, Accountability, Privacy, Security, and Resilience.

BAGI serves a function similar to a credit score, delivering insurers a measurable risk assessment tool. In coming years, insurers will likely inquire, "What is your BAGI score?"

The ramifications of BAGI on underwriting are groundbreaking. For example, a multifaceted real estate investment trust (REIT) with a BAGI score of 72 (indicative of adequate governance) poses less risk than one with a score of 31 (reflecting minimal oversight). Furthermore, a score of 89 evidence comprehensive governance improvements, likely resulting in favorable premium terms.

Predictions for premium impacts across portfolios extending to 500 million square feet are as follows:

• BAGI score 31-50 (minimal governance): +2.0% premium increase due to governance risk.

• BAGI score 51-75 (partial governance): +0.5% premium increase, with room for reductions as scores rise.

• BAGI score 76+ (comprehensive governance): No premium adjustment with the potential for price breaks.

For a portfolio valued at $500 million currently spending $2.5 million annually on insurance, a mere 1% reduction translates to $25,000 in yearly savings, underscoring the economic benefits of investing in stronger governance frameworks.

---

Navigating Regulatory Pressures

Three regulatory trends are converging to underscore the urgency of AI governance discussion:

1. The EU AI Act: Upcoming phased enforcement enacts stringent requirements for high-risk systems, including building safety measures. US REITs with European interests will face compliance pressures, leaving insurers vulnerable to unprepared liability challenges.

2. NYC LL97 and Building Performance Standards: New York City’s Local Law 97 mandates measurable emission reductions by 2030, relying on automation to ensure compliance. Regulators will require verifiable proof of AI governance, impacting insurers responsible for these properties.

3. State-Level AI Liability Frameworks: Emerging legislation, such as Colorado's AI Act, introduces explicit liabilities associated with AI failures, positioning autonomous systems lacking governance protocols as potentially dangerous—prompting urgent changes in oversight.

As regulatory landscapes evolve, the insurance sector must adapt pricing mechanisms to reflect changing liability exposures, setting the stage for insurers to transition toward proactive governance requirements.

---

The Economic Argument: Governance Equals Lower Premiums

Consider the following scenarios illustrating the financial advantages of implementing strong governance frameworks:

Scenario 1: Mid-Sized Office REIT (200 buildings, 15M sq ft)

• Current annual premium: $3.2M

• AI governance risk exposure (ungoverned agents): +$160K/year (5% increase)

• Governance implementation cost: $240K (one-time)

• Implementation ROI: 1.5 years (recovered through premium reduction)

• Annual savings: $160K (with potential for growth)

Scenario 2: Large Industrial/Data Center (45 buildings, 50M sq ft)

• Current annual premium: $12.8M

• AI governance risk exposure (ungoverned agents): +$1.28M/year (10% increase)

• Governance implementation cost: $850K (one-time)

• Implementation ROI: 0.67 years (recovered in 8 months)

• Annual savings: $1.28M (and increasing)

Scenario 3: Healthcare System (28 buildings, 8M sq ft)

• Current annual premium: $5.6M

• AI governance risk exposure: +$840K/year (15% increase)

• Governance implementation cost: $380K (one-time)

• Implementation ROI: 0.45 years (recovered in 5 months)

• Annual savings: $840K (and compounding)

These examples substantiate the financial advantages of governance investment, with recent trends indicating that as insurers clearly define and price AI governance risk, the disparities in premium rates are set to burgeon.

---

Actionable Steps

For Insurance Carriers: Now is the time to initiate action. Establish a governance risk assessment framework. Evaluate your top 20 accounts, conducting governance gap assessments. Identify facilities using autonomous agents lacking adequate governance and adapt underwriting models accordingly to maintain competitive advantage.

For Building Operators or REITs: Conduct a governance gap assessment of your systems deploying autonomous agents. Gain clarity on which systems are well-governed versus those needing oversight. Develop your governance framework proactively, recognizing that implementation timelines can be condensed to four to eight weeks on a per-building basis, with substantial long-term returns.

For Brokers: Facilitate discussions with clients about governance. Inquire into the role of autonomous agents within their facilities, ask about operational constraints, and request evidence validating the AI's decision-making process. The answers provided will become integral to insurance coverage discussions.

Historically, the insurance sector has adapted to address emerging cyber risks; now the industry must recognize the imminent necessity for AI governance in buildings. As this domain progresses towards regulatory and underwriting imperatives, it raises the stakes for all stakeholders involved in building operations.

---

About the Author

James C. Waddell is President of Cognitive Corp and the guiding force behind the Building Constitution and BAGI frameworks. His extensive research, comprising over 175 papers on AI governance in the built environment, represents a comprehensive look into effective management practices for governing autonomous systems.

---

Cognitive Corp

AI Governance for Building Operations

[hello@cognitivecorp.com](mailto:hello@cognitivecorp.com)

www.cognitivecorp.com

*Building Constitution is a registered trademark of Cognitive Corp. BAGI (Building AI Governance Index) is a proprietary assessment framework developed by Cognitive Corp.*

---

Keywords: AI governance, autonomous agents, Building Constitution, insurance industry, building operations, risk management, BAGI, smart buildings, regulatory compliance, insurance premiums, decision making, accountability, risk assessment