Smar NetworkSmar.inSmargrid.comNewsInsuranceRoomsGamesMeasurements

SMAR Daily Industry News

Daily news across AI, technology, finance, insurance, healthcare, automotive, real estate, and government

Facebook Telegram
← Back to Insurance News

Insurance AI Success Depends on Operational Discipline

📅 2026-11-12 | 📰 Insurance Business America
#insurance ai #operational discipline #resource pro #ai adoption #insurance technology

The insurance industry has utilized artificial intelligence tools for years, yet the organizations achieving the most success are not necessarily those with the most advanced technology. According to Michael Costonis of ReSource Pro, the carriers and managing general agents (MGAs) pulling ahead are those that prioritized operational order before scaling their AI initiatives.

Costonis emphasizes that true value from AI is derived from discipline, transparency, guardrails, and thoughtful execution. Without these foundational elements, AI implementation often results in fragmented, inefficient operations. Organizations that lack structured processes and available data struggle to harness the technology's full potential. In these environments, AI usage is often relegated to isolated pockets where individuals use tools for productivity hacks, but the work is frequently pushed upstream unnecessarily, making it difficult to measure actual performance improvements.

A common pitfall occurs when companies introduce AI solutions without a standardized approach. When AI is pushed to individual users without oversight, employees often create their own unique methods for handling tasks. This leads to inconsistencies across different offices, functions, and teams. Such environments are characterized by redundant workflows, mismatched data, and scaling inefficiencies. Ultimately, this lack of discipline prevents organizations from trusting AI-generated results, obscures transparency regarding how those results were produced, and increases the workload for end users.

To move beyond this, Costonis advocates for managed agentic services, which involve humans working alongside machines to refine workflows and provide feedback to the model. This collaborative approach allows for adjustments to procedures and business structures, which in turn improves the model's performance. He notes that effective AI deployment requires a combination of human oversight and machine capability.

Costonis categorizes insurance organizations into two groups: experimenters and those taking a deliberate approach. Experimenters have rolled out AI tools to staff for individual tasks, but these efforts often fail to scale or achieve broad adoption. Conversely, organizations capturing real value examine entire functions from end to end. They verify that processes are standardized, data is accessible, and they make clear decisions about when to use AI versus other techniques like deterministic code.

The litmus test for proper AI implementation is the ability to maintain efficiency throughout an entire workflow. If an AI tool produces an output, but that output requires manual evaluation by an underwriter or supervisor, the efficiency gains are negated. Leaders should evaluate their end-to-end workflows to identify where value is created and where it is destroyed, ensuring that the resulting decisions are trustworthy and transparent.

Why it matters

For insurance leaders, this highlights that operational maturity and standardized processes are prerequisites for successful AI adoption, rather than the technology itself.

FAQ

What are the key components required to power AI effectively?
According to Michael Costonis, you need structured processes, available data, clear decision rights regarding how and when decisions are made, and a defined outcome.

What is the primary risk of deploying AI without operational discipline?
It leads to inconsistent processes across functions, data mismatches, redundancy, and a lack of trust in the results, which ultimately pushes more work back onto the end user.

How can leaders tell if their organization is using AI correctly?
Leaders should look at the end-to-end workflow to see if the AI actually creates efficiency or if it requires manual intervention by supervisors, which destroys the value of the tool.

Source: Insurance Business America. See the original source for full details.