Regulatory Compliance

Insurers navigate data and talent hurdles to unlock AI potential in underwriting and claims

Wednesday, 10 December 2025 7:42AM UTC

Major challenges around data ownership, quality, legacy systems, and workforce shortages are delaying the adoption of AI across the insurance industry. Experts emphasise the importance of robust governance, phased modernisation, and strategic talent development to harness AI’s full potential while safeguarding privacy and compliance.

Data lies at the heart of every insurance artificial intelligence deployment, but ambiguity over who owns which datasets, how they can be used, and whether they are sufficiently accurate and accessible is hobbling many programmes before they begin. According to the original report, insurers draw data from policyholders, providers, vendors and public bodies , each with different legal and contractual constraints , leaving organisations unsure who may lawfully reuse information for model training, analytics or automation. ^[1]^[4]

Beyond ownership, poor data quality and fragmentation are pervasive. Industry analysis shows many insurers hold customer, policy and claims records in isolated silos and legacy systems that lack standardisation; incomplete, inconsistent or poorly labelled records produce unreliable AI outputs and increase the chance of erroneous underwriting, pricing or claims decisions. Building master data pipelines and clear data stewardship is therefore a precondition for trustworthy AI. ^[1]^[2]^[3]

Legacy technology is a practical blocker to integration. Older billing, policy administration and claims platforms were not designed to interoperate with modern machine‑learning tools; the cost and complexity of refactoring or replacing such systems is cited repeatedly by insurers as a primary constraint on delivery. Connecting legacy stacks through APIs or middleware, and adopting modular, scalable components, is the pragmatic path many organisations are taking. ^[1]^[2]^[4]

Strong governance is essential to bridge legal, ethical and operational gaps. The company said in its announcement that policies governing collection, access, retention and sharing must reflect sector laws , notably HIPAA for health data , and be coupled with technical controls. Researchers and regulators alike caution that weak governance raises the risk of privacy breaches, regulatory non‑compliance and reputational damage as generative and other advanced models are rolled out. ^[1]^[6]^[7]

Healthcare‑adjacent insurers and medical practice administrators face heightened sensitivities because patient information attracts special protections. The lead report highlights approaches such as federated learning and privacy‑preserving analytics that allow models to benefit from distributed datasets without centralising identifiable patient records , an option that can reduce legal exposure while enabling fraud detection and clinical risk modelling. ^[1]

Workforce shortages compound technical obstacles. The sector struggles to recruit data scientists, ML engineers and AI product specialists who understand both advanced techniques and insurance operations; surveys also show a generational expectation gap, with younger employees keen to use AI but feeling unsupported by employers. Insurers report budget and resource limits and are increasingly prioritising data literacy as a strategic gap to close. ^[1]^[5]^[4]

To offset talent constraints, many firms are turning to managed AI services and no‑code platforms. Outsourced providers deliver pre‑built models, operational support and infrastructure, enabling medium and smaller insurers , and medical practices with limited IT headcount , to adopt automation without building large in‑house teams. The trade‑off is the need for careful vendor selection and contract governance to preserve data controls and regulatory accountability. ^[1]^[4]

When properly governed and integrated, automation delivers measurable operational gains. Optical character recognition, NLP and rules‑augmented ML streamline claims intake, surface suspicious patterns for fraud detection, and improve customer interactions through conversational agents , reducing processing times and manual error rates. Yet industry data warns that algorithmic bias, model “hallucinations” and adversarial risks mean automated decisions require explainability, audit trails and human oversight. ^[1]^[2]^[3]^[7]

Successful AI adoption is as much cultural and leadership work as it is technical. Insurers that align AI projects with corporate strategy, fund training, and embed ethics and compliance into programme governance accelerate deployment and protect customers. Modernisation efforts must therefore pair IT roadmaps with change management that equips operational teams , from medical administrators to claims handlers , to use AI effectively. ^[1]^[4]^[7]

Looking ahead, advances in natural language processing, predictive analytics and AI‑enhanced cybersecurity will increase the value of properly governed, high‑quality data and deepen automation across underwriting, care coordination and fraud prevention. For medical practice owners and IT managers, the practical checklist is consistent: shore up data governance, plan phased modernisation, invest in skills and partnerships, and insist on transparency and controls in every vendor relationship. Those steps will determine whether AI becomes a productivity and care‑quality lever rather than a compliance and reputational risk. ^[1]^[7]

##Reference Map:

^[1] (Simbo.ai blog) - Paragraph 1, Paragraph 2, Paragraph 3, Paragraph 4, Paragraph 5, Paragraph 6, Paragraph 7, Paragraph 8, Paragraph 9, Paragraph 10
^[2] (Tribe.ai) - Paragraph 2, Paragraph 3, Paragraph 8
^[3] (Invensis) - Paragraph 2, Paragraph 8
^[4] (BusinessWire report) - Paragraph 1, Paragraph 3, Paragraph 6, Paragraph 7, Paragraph 9
^[5] (Insurance Business magazine) - Paragraph 6
^[6] (PR Newswire) - Paragraph 4
^[7] (Deloitte) - Paragraph 4, Paragraph 8, Paragraph 9, Paragraph 10

Source: Noah Wire Services

More on this

https://www.simbo.ai/blog/challenges-in-ai-implementation-within-the-insurance-sector-overcoming-data-ownership-and-talent-shortages-2217953/ - Please view link - unable to able to access data
https://www.tribe.ai/applied-ai/challenges-of-applying-ai-in-insurance - This article discusses the challenges insurers face when integrating AI, particularly focusing on the integration with legacy systems and the quality and availability of data. It highlights that outdated IT systems often lack compatibility with new AI solutions, making integration difficult and costly. Additionally, the article points out that AI models require large, accurate, and well-structured datasets, and if the data is incomplete or inconsistent, it can lead to inaccurate predictions and errors in risk assessment, pricing, and claims processing.
https://www.invensis.net/blog/ai-in-insurance - This article outlines five key challenges in implementing AI within the insurance sector: isolated, low-quality data; integration with legacy systems; data privacy and regulatory compliance; algorithmic bias and discrimination; and transparency and accountability. It emphasizes the importance of investing in high-quality data and strong governance to address these challenges, suggesting that insurers should build robust data pipelines, employ master data management tools, and use explainable AI to understand decisions.
https://www.businesswire.com/news/home/20250904625069/en/Insurers-Say-Data-AI-Central-to-Growth-but-Gaps-in-Strategy-Ownership-and-Delivery-Remain - This report highlights that while insurers recognize the importance of data and AI for growth, there are significant gaps in strategy ownership and delivery. It reveals that only 11% of insurers are prioritizing data literacy, and 29% are enabling data-driven decision-making. The report also notes that 49% of insurers cite budget or resource constraints as a top blocker, with smaller firms facing resourcing gaps and larger firms struggling with legacy systems and data quality and governance.
https://www.insurancebusinessmag.com/us/news/breaking-news/gen-z-insurance-workers-say-industry-risks-falling-behind-on-ai-survey-550733.aspx - This survey reveals that while Gen Z values the stability and career prospects of the insurance sector, many believe the industry is lagging in technological adoption. Nearly 70% of Gen Z respondents believe AI could improve their workflow, yet only 8.5% reported being strongly encouraged by their employers to use AI. The survey also highlights that 45% of respondents cited slow technology uptake as a serious problem for the sector.
https://www.prnewswire.com/news-releases/insurers-urged-to-safeguard-data-amid-rising-ai-adoption-and-privacy-risks-says-info-tech-research-group-302331530.html - This article discusses the risks associated with generative AI in the insurance industry, including data breaches of personally identifiable information (PII), noncompliance with regulations like GDPR and HIPAA, and insider threats. It advises insurers to implement robust data governance practices, ensure transparency, and foster customer trust in the responsible use of AI. The article also emphasizes the importance of identifying insurance-specific risks and adopting a continuous improvement approach supported by metrics and a risk-based strategy aligned with a privacy framework tailored to organizational needs.
https://www2.deloitte.com/us/en/insights/industry/financial-services/scaling-gen-ai-insurance.html - This article discusses the risks and regulations associated with generative AI in the insurance sector, including algorithmic bias and discrimination, the phenomenon of 'hallucinations' where AI systems generate incorrect or misleading information, and the complexity of integrating multiple internal and external data sets, which can increase vulnerability to cyberattacks. It emphasizes the need for insurers to incorporate risk management and governance into their AI scaling efforts from the start.

Noah Fact Check Pro

The draft above was created using the information available at the time the story first emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed below. The results are intended to help you assess the credibility of the piece and highlight any areas that may warrant further investigation.

Freshness check

Score: 8

Notes: The narrative appears to be original, with no evidence of prior publication. The earliest known publication date is 2025-09-04. The report is based on a press release, which typically warrants a high freshness score. No discrepancies in figures, dates, or quotes were found.

Quotes check

Score: 10

Notes: No direct quotes were identified in the narrative, indicating potential originality or exclusivity.

Source reliability

Score: 7

Notes: The narrative originates from Simbo AI, a company with a limited online presence. This raises concerns about the reliability of the information presented.

Plausibility check

Score: 9

Notes: The claims made in the narrative align with known challenges in AI implementation within the insurance sector, such as data ownership issues and talent shortages. However, the lack of external verification from reputable sources reduces the overall confidence in the narrative's accuracy.

Overall assessment

Verdict (FAIL, OPEN, PASS): OPEN

Confidence (LOW, MEDIUM, HIGH): MEDIUM

Summary: The narrative presents plausible challenges in AI implementation within the insurance sector, with no evidence of recycled content or discrepancies. However, the limited online presence of Simbo AI and the absence of direct quotes or external verification from reputable sources raise concerns about the reliability and accuracy of the information presented.

AI
insurance
data governance