Insurance underwriting is moving into a new era, one shaped not just by changing risks, but by the technology we use to evaluate them. Artificial intelligence (AI) is changing underwriting quicker than many systems and plans can keep up. Leaders in insurance will succeed not by using AI simply because it exists, but by integrating it with clear goals, strong rules and forward-thinking strategies.
According to
Moving from reactive to predictive underwriting
Historically, underwriting has been a slow, manual process reliant on historical data and broad risk proxies. AI is changing that dynamic dramatically.
Today, underwriting systems can synthesize application data, claims history, public records, IoT device outputs and satellite imagery — all within minutes. Even more important, AI allows insurers to dynamically model risk based on real-time inputs like driving behavior, wearable health data and climate analytics.
The result is not just faster decision-making,
AI: A new frontier for prevention and a new source of exposure
AI is also redefining the relationship between underwriting and risk prevention.
Connected devices can identify property risks before claims occur. Wearables promote healthier behavior, tying premiums to proactive risk mitigation. AI-driven monitoring tools help businesses manage operational exposures before they escalate.
Yet every application of AI introduces its own set of new risks — for both insurers and their clients.
Consider the possibilities:
· Product liability claims stemming from AI-generated designs that malfunction.
· Employment discrimination lawsuits triggered by AI-enabled hiring tools that introduce biases.
· Cyber breaches caused by AI-driven phishing attacks or deepfakes.
· Errors and omissions exposures when AI tools synthesize inaccurate data that businesses then rely on.
Many organizations are adopting AI tools quickly without fully considering how their use could expose them to financial losses, liability claims or reputational harm. As insurers, we have a dual obligation: to help our clients understand these new risks while managing them internally.
Balancing speed and trust in AI adoption
Bias and transparency remain critical concerns. Even when datasets appear neutral, AI models can replicate systemic inequities. Underwriting decisions powered by AI must be explainable, auditable and defensible — not just to internal stakeholders, but increasingly to regulators and courts.
Transparency cannot be optional. Trust depends on it.
Meanwhile, infrastructure challenges persist. Many insurance carriers still operate on siloed systems, hampering the effective use of AI. Feeding fragmented or inconsistent data into AI models only amplifies inaccuracies. True digital transformation requires a connected, consistent infrastructure — and investment in training people to critically assess AI outputs, not just deploy them.
Underwriting AI — and insuring it
As AI becomes integral to underwriting operations, it is simultaneously becoming a new source of insurable risk.
Yet, standard commercial insurance policies rarely address AI-specific exposures today. While some carriers have introduced endorsements that expand cyber coverage definitions to include AI-related incidents, these solutions primarily address a narrow set of risks. Broader exposures — spanning general liability, errors and omissions, and directors and officers — remain uncharted territory.
History suggests the insurance market will eventually respond — often by excluding unquantifiable exposures. We've seen it happen with infectious disease and biometrics. AI could follow a similar path, prompting the need for specific endorsements or new policy structures.
Organizations that fail to develop AI governance frameworks and policies now may find themselves uninsured against significant future losses.
Focusing on future-ready underwriting
The future of underwriting will not be driven by chasing the latest AI model. It will be shaped by asking better questions and leading with intention.
· Are our AI models explainable, auditable and defensible?
· Are we bridging silos across underwriting, claims and service to create connected strategies?
· Are we proactively training our people to evaluate and govern AI outputs?
· Are we using client and third-party data responsibly, with fairness at the forefront?
Insurance companies that thrive in this AI-driven future will not just deploy the latest models. They will lead with governance, transparency, and a relentless focus on their core mission: helping clients navigate risk with clarity and confidence.
The opportunity is here — and it belongs to those who act with both speed and intention.
