July 2025

AI in Underwriting: How It’s Transforming the Insurance Industry

Jay D'Aprile

Executive Vice President

The core promise of AI and automation is to make tasks simpler, easier, and faster. For the insurance industry, especially AI in underwriting, growing complexities makes this need even more urgent.

Rapid regulatory changes, geopolitical and macroeconomic uncertainty, ecosystem integration and embedded insurance—plus the explosion of data associated with AI, IoT, and connected devices—all introduce new risks. AI can help underwriters keep apace with these changes, but only if they take advantage of it. If not, these professionals won’t be standing still; they’ll be falling behind.

AI and Automation in Underwriting

AI and automation in underwriting help streamline, enhance, and even transform the traditional insurance underwriting process:

Automating data collection and extraction through use of natural language processing (NLP) and other machine learning (ML) techniques, reducing the need for manual data entry
Deploying ML models to assess risk more accurately by looking at vast datasets vs. the same four or five categories
Accelerating repetitive, low-value tasks through automated workflows—downloading documents, pre-filling forms, triaging submissions, sending routine communications, etc.
Leveraging generative AI (GenAI) models to review and interpret complex documents, compare extracted data against underwriting rules, and provide more transparent justifications for decisions

The AI Revolution in Underwriting

The AI revolution in underwriting may seem like a sudden shock, but it has been a long time coming. The need for simpler, faster, more accurate processes is more poignant than ever, as longer claims cycles lead to decreased customer satisfaction, which in turn lead to customers seeking alternative insurance providers. On top of that, the growing complexity of cyber and climate risks makes traditional risk assessment practices less accurate than before.

AI in underwriting promises to improve both speed and accuracy, which means a significant lift in business value. A recent study showed that AI reduces the average underwriting decision time for standard policies from three to five days down to just 12.4 minutes. And it’s done all of that by maintaining a 99.3% accuracy rate in risk assessment. For complex policies, the gains are more modest but still impressive: 31% reduction in processing time, while improving accuracy by 43%.

Human + Machine: A Hybrid Future

Given these impressive results, is it still worth it to hire human underwriters? Could organizations simply let some of these AI solutions handle the majority of cases?

The reality isn’t so straightforward. Consumer trust in insurance companies remains at a low 43%, while 42% of consumers are concerned with AI making mistakes in its predictions and negatively impacting their coverage. Fully automating the underwriting process will not help close that trust gap.

However, a hybrid approach that balances AI-powered efficiency gains with professional judgement and expertise—a “human-in-the-loop” approach—can free underwriters to spend more time on high-complexity risk assessments, strategic input, and building relationships with their customers. This approach offers the best of both worlds, with humans and AIs both leaning into their strengths.

The Bottom Line: People Will Never Go Out of Business

Although AI promises to redefine the risk landscape, the value of human intelligence is still high. The truth is, every claim is different, and every situation has a unique set of factors that influence its risk level. Data and predictive analytics can provide insights, but human judgment, built from years and decades of experience, is still critical to make the best decision for all parties involved.

Talent Challenges & Opportunity Gaps

Given this hybrid future, how should insurance and financial services leaders think about hiring underwriters? First, there’s a growing demand for AI and data-savvy underwriters, with the global market for AI underwriting is projected to hit $46B by 2031. What’s more, many skilled underwriters are set to retire in the coming decade, further exacerbating the demand for talent.

However, talent supply is not currently large enough to meet this demand. Only one-third of insurers have formal AI training programs in place for their staff, while fewer than 20% of underwriters are considered “AI-savvy.”

However, adopting AI without simultaneously investing in the talent needed to manage it presents its own set of risks. One example is that of bias. Some regulators warn that hyper-personalization could lead to the creation of “uninsurable” classes, which exacerbates biases against historically underrepresented groups. A human underwriter who understands the inner workings of AI can help to identify and curtail these risks as they arise.

Other risks associated with AI implementation sans qualified talent include:

Unrecognized errors and outdated information that reduces the data quality, leading to inaccurate assessments
No flexibility for edge cases and outliers
Cybersecurity and data privacy risks, as AIs process massive amounts of sensitive data
Loss of customer trust and insurer credibility when changes happen and there is no human person to explain them

For insurance and financial services leaders, this presents a major risk that, if unaddressed, could lead to severe talent shortages and a corresponding decline in performance. Although insurance is a typically conservative and traditional sector, leaders must embrace change or risk falling behind.

Implementing AI Responsibly

AI has the potential to transform many functions within the insurance industry, leading to greater efficiencies and scale. However, realizing those benefits will not happen automatically. True value creation is only possible when insurance leaders embrace both AI and the need for responsible human oversight.

Here are some strategies that may benefit insurance and financial services executives:

Deploying cross-functional teams—integrating tech, underwriting, talent, and compliance—for a more holistic approach
Shifting recruiting and talent search priorities to include skills like AI fluency, data literacy, change leadership, etc.
Shifting internal cultures to embrace continuous learning and technical collaboration
Positioning employer brands as innovators to attract top-tier talent

Final Thoughts on AI in Underwriting

AI is transforming underwriting, no question about it. However, top-tier talent is the force multiplier. By 2035, half the workforce will be retiring, leaving close to 400,000 roles unfilled. Without skilled people to fill those roles, AI could become a net-negative.

Rather than avoid or put off AI, now is the time to embrace it. Leaders should start focusing their talent strategies on attracting tech-savvy underwriters and plan their workforce with the future in mind.