🚀 New Holistic AI Tracker 2.0 – Now featuring Atlas, a heat-mapped world view tracking AI legislation, incidents, and regulations in real time.
Register for free
Learn more about EU AI Act
Join Webinar: Bias Detection in Large Language Models - Techniques and Best Practices
Register Now
Learn more about EU AI Act

Holistic AI’s Response to Colorado’s Bias Audit Proposals for Underwriting in Life Insurance

Authored by
No items found.
Published on
Oct 20, 2023
share this
Holistic AI’s Response to Colorado’s Bias Audit Proposals for Underwriting in Life Insurance

Holistic AI has submitted written comments to Colorado’s Division of Insurance in response to a draft algorithm and predictive model quantitative testing regulation for underwriting in life insurance as part of the consultation process for SB169. This follows the announcement of the enforcement of Governance and Risk Management Framework Requirements for Life Insurers due to be enforced from 14 November 2023.

Introduced to prohibit unfair discrimination from the use of external consumer data and information sources, algorithms, and predictive models in insurance practices in Colorado, SB169 was passed in the middle of 2021 and due to be enforced from 1 January 2023 at the earliest.

The draft regulation requires insurers to carry out annual testing to determine whether data sources, algorithms, or predictive models result in unfair discrimination based on inferred race/ethnicity using logistic regression for eligibility determinations and linear regression for insurance premiums.

Key takeaways from Holistic AI’s comments

While the quantitative assessment, or internal bias audit, is a step in the right direction to ensuring that insurance practices do not unfairly discriminate, Holistic AI raised points of contention around:

  • The use of tools to infer race/ethnicity instead of relying on self-reported data
  • The focus solely on race/ethnicity, particularly since the law prohibits discrimination based on race, color, national or ethnic origin, religion, sex, sexualorientation, disability, gender identity, or gender expression
  • Using white insureds as a group of reference instead of the group with the highest rate since other group differences may be missed
  • The need for independent testing to ensure the robustness of the quantitative assessments
"Governance doesn’t slow us down; it accelerates AI adoption, ensures compliance, and boosts our ability to realize digital goals."
- Sam Dover, Global AI Assurance & Strategy Lead

Unilever’s Success Story:

  • Over 300 AI use cases deployed across their operations, from marketing to supply chain optimization.
  • The use of AI assurance frameworks has ensured alignment with global regulations, including the EU AI Act.
  • A cross-functional AI Center of Excellence in London supports scalability and innovation.

Key Highlights from the Talks

1. Addressing Challenges in AI Deployment

  • 80% of AI projects fail to scale, and 50% of organizations cite risks and lack of processes as the main barriers.
  • Key challenges include:
  • Lack of a clear ROI realization plan.
  • Absence of frameworks for testing and reporting.
  • Gaps in end-user trust and understanding of AI systems.
  • Regulatory hurdles such as the EU AI Act and over 200 state-level AI bills in the U.S.

2. Proven Business Benefits of AI Governance

The adoption of AI governance platforms delivers measurable impact:

  • 50–60% faster AI deployment into production, unlocking business value.
  • 25–40% reduction in wasted spend by proactively identifying and addressing risks.
  • Increased customer trust ratings and regulatory compliance scores by 30%.
Adriano Koshiyama, Co-Founder of Holistic AI mentioned that:
"Without structured governance, businesses face mounting barriers, from trust issues to regulatory fines. Governance transforms AI into a scalable, impactful tool."

3. Solutions to Accelerate AI Adoption

Holistic AI’s Governance Platform provides:

  • A management hub offering transparency and control over both internal and third-party AI systems.
  • A focus on risk mitigation, addressing issues like bias, privacy, and security.
  • Tools to establish a common framework for testing and ensuring compliance across the organization.
As Adriano Koshiyama concluded:
"By embedding governance, businesses can reduce ethical incidents by 40% and position themselves as leaders in AI innovation."

Key Takeaways for Businesses

Establish an AI Center of Excellence

Ensure alignment with business goals and foster an AI-driven mindset

Invest in Governance Early

Build trust, minimize risks, and accelerate ROI on AI investments

Govern Across the Full AI Lifecycle

Leverage purpose-built solutions designed for the complexity and scale of AI

Download our comments here

Videos sourced from Microsoft Ignite 2024.

DISCLAIMER: This news article is for informational purposes only. This blog article is not intended to, and does not, provide legal advice or a legal opinion. It is not a do-it-yourself guide to resolving legal issues or handling litigation. This blog article is not a substitute for experienced legal counsel and does not provide legal advice regarding any situation or employer.

See the industry-leading AI governance platform in action

Schedule a call with one of our experts

Get a demo