Motion Consulting Group · Kelly
AI Governance
Prepared for BCBS of Illinois
Confidential draft for review. Please enter the access code.
Invalid code.
AI Governance
Initial point of view for discussion · prepared by Motion Consulting Group (a Kelly company) for Blue Cross and Blue Shield of Illinois
AI Governance · Prepared for Blue Cross and Blue Shield of Illinois

Govern AI with confidence and care.

BCBSIL is putting structure and accountability around AI before scaling it across the enterprise — the right instinct for a health plan millions of members trust. Here's how Motion Consulting Group helps you build that foundation: design the operating model, implement the controls, and run the system of record that keeps AI safe, compliant, and worthy of that trust.

An initial point of view from Motion Consulting Group, a Kelly Services company.
01 · Why Now

Govern AI before you scale it.

AI is no longer an experiment at BCBSIL — it's a strategic capability that touches members, care, and claims. The leaders who scale it well put the rules, ownership, and accountability in place first. That's exactly the foundation this engagement builds.

You've already laid the groundwork — GitHub and Copilot are in place, and teams are using AI today. The question now isn't whether to use AI; it's how to govern it so adoption scales safely. For a health plan, that bar is high: AI decisions can affect a member's care and coverage, which means governance has to hold up to HIPAA, the NAIC Model Bulletin on AI that states are adopting, model-risk scrutiny, and — above all — member trust.

Get this right and governance isn't a brake on innovation. It's what lets you move faster with confidence: protecting members, earning trust, and reducing risk, all at once.

75%
of the world's economies expected to be covered by AI regulation by 2030, with rules multiplying (Gartner)
growth in board-level AI-risk oversight in a single year, 2024→2025 (EY, Fortune 100)
Now
the NAIC AI Model Bulletin is being adopted state by state — insurers are expected to govern AI use

The member-first frame: every control in this program traces back to a simple promise — that AI at BCBSIL is used responsibly, explainably, and in the member's interest.

02 · The Engagement

Assess. Design. Implement. Operate.

A clear, four-phase path — and the difference is the back half. Many firms will hand you a policy and a gap assessment. MCG designs the operating model and then builds and runs it — the part that turns governance from a document into a working system.

Phase 1

Assess

Inventory your AI estate, classify risk, map regulatory exposure, and benchmark against your target framework.

4–12 weeks
Phase 2

Design

Operating model and decision rights, AI policy, the use-case intake and review process, and your risk taxonomy.

2–6 weeks
Phase 3

Implement

Stand up the model registry, intake gate, controls, and Copilot guardrails; integrate with your existing GRC. Where MCG builds.

weeks–months
Phase 4

Operate

Continuous monitoring for drift and risk, board reporting cadence, and audit readiness — run as a managed service. Where MCG stays.

ongoing

Start small, prove it, scale it. The assessment opens the engagement and gives you a concrete, prioritized picture — not a year-long commitment before you see value.

03 · Copilot Governance

Make the most of the investment you've already made.

GitHub Copilot is already in your environment. The fastest, most tangible win is governing it well — so developers move faster and the organization stays protected.

Usage standards & guardrails

Clear, practical rules for what Copilot can and can't touch — protected data, sensitive repos, approved use patterns — so teams have confidence, not guesswork.

Secure-SDLC controls

Code-generation risk addressed where it lives — IP and license hygiene, secret-scanning, security review, and quality gates wired into your pipeline.

Developer enablement

Adoption playbooks and training so the benefit is realized broadly and consistently — not just by the early adopters.

Measured value

Track adoption, productivity, and cost so the Copilot investment is governed and demonstrably paying off.

04 · Operating Model

What you're left running.

Governance only works if it operates day to day. This engagement stands up a working model that plugs into the risk infrastructure BCBSIL already has — it doesn't bolt a parallel process beside it.

  • A joint review body — the right people (security, privacy, compliance, data, and the business) reviewing production-intent AI, with clear, documented decision rights.
  • An intake gate — a single front door where every new use case is described, risk-classified, and approved before it reaches members or production.
  • A model registry as the system of record — every AI system inventoried, tiered, owned, and tracked across its lifecycle.
  • Bias & fairness review — disparate-impact testing wherever AI informs care, coverage, or member-facing decisions, so health equity is governed, not assumed.
  • GRC integration — because the standard we build on aligns to your existing risk management, governance slots in rather than standing alone.
  • An audit and board cadence — what's measured, reviewed, re-certified, and reported, on a rhythm leadership can rely on.
05 · Framework Alignment

Built on the standards your regulators and auditors expect.

These frameworks stack rather than compete. We use ISO/IEC 42001 as the management-system backbone and NIST AI RMF as the risk method, then layer the cybersecurity substrate, the AI-specific threat surface, and a healthcare-AI overlay on top — and map every deliverable to the rule it satisfies.

FrameworkWhat it isWhy it matters to BCBSIL
ISO/IEC 42001Certifiable AI management system; aligns to ISO 27001/9001The governance backbone — and it plugs into your existing risk stack
NIST AI RMFGovern / Map / Measure / Manage risk methodA defensible, widely-recognized way to measure and manage AI risk
HIPAAProtected health information safeguardsNon-negotiable wherever AI touches member health data
NAIC AI Model BulletinInsurer expectations for governing AI use, adopted state by stateDirectly on point for a health plan — be ready as states adopt it
Model-risk disciplineValidation, monitoring, and documentation of consequential modelsApplies the rigor regulators already expect to your AI estate
NIST CSF 2.0 · CIS Controls v8.1Current cybersecurity framework and controls baselineThe security posture your AI inherits — governed to current versions, not last decade's
OWASP LLM Top 10 · MITRE ATLASThe AI-specific attack surface — prompt injection, model abuse, agentic riskWhat a security-only review misses — real wherever AI faces members or sensitive data
The Joint Commission · CHAI model cardsResponsible-AI guidance for health + the model "nutrition label" for health AISpeaks a health plan's world — accreditation-aligned and member-facing, not generic

One crosswalk, not silos. Instead of a separate binder per regulation, you get a single map: each control, written once, traced to every standard it satisfies — far easier to operate and to defend in an audit.

06 · Standards & Enablement

From ad-hoc use to a governed practice.

As AI use spreads, consistency becomes the risk. We help BCBSIL turn scattered, individual usage into a repeatable, governed practice the whole organization can rely on.

Prompt engineering & usage standards. Practical standards for how AI is prompted, reviewed, and used safely — especially where outputs touch members or sensitive data — so quality and safety don't depend on who happens to be at the keyboard.

An AI Center of Excellence. A home for patterns, reusable controls, training, and shared learning — so good practice spreads and every team isn't solving the same governance problem from scratch.

07 · Readiness & Roadmap

Know where you stand. See the path to scale.

Before you build, you get a clear, honest read on where BCBSIL is today — and a sequenced roadmap for getting to enterprise-scale, responsible AI.

Readiness assessment

A grounded view of your AI estate, governance maturity, regulatory exposure, and the gaps that matter most — prioritized, not a laundry list.

Enterprise roadmap

A phased plan that sequences the highest-value, lowest-risk moves first, with change management so the framework is adopted — not shelved.

08 · Why Motion Consulting Group

Governance designed by people who build and run AI.

Advisory firms can hand you a framework. MCG designs it, implements it, and operates it — backed by Kelly's scale and talent bench. For a regulated enterprise that has to live with this every day, that's the difference that matters.

Advisory-only firms

  • Assess and recommend a framework
  • Hand over a policy and a gap report
  • Strong on design — then the work goes back to your teams

Motion Consulting Group

  • Design the operating model and build it
  • Stand up the registry, intake, controls, and monitoring
  • Run it as a managed service — engineering + Kelly's talent at scale

Independence where it counts. The team that builds and runs your controls is firewalled from the party that attests to them — so a build-and-run model still produces an assurance your auditors, board, and regulators can rely on.

Proof points (representative — final, cleared references to be confirmed): AI governance and delivery across regulated, high-stakes environments — healthcare (the closest analog for BCBSIL), a top-tier US telecommunications carrier, energy, and biotech.

09 · The Path Forward

Let's start with a focused assessment.

The simplest first step is a scoped Phase 1 assessment: a clear picture of your AI estate, your governance gaps, and a prioritized plan — delivered quickly, with no long commitment to begin.

Let's scope a working session.

A short session with your security, privacy, compliance, and data leaders to walk this framework, hear your priorities, and shape a Phase 1 assessment tailored to BCBSIL.

Working session — scheduled through your MCG contact
Initial draft for discussion. This is an early point of view prepared by Motion Consulting Group (a Kelly Services company) for Blue Cross and Blue Shield of Illinois. It is not a proposal, quote, or commitment, and is intended to frame a conversation. Market figures are drawn from public 2025–2026 research (Gartner, EY, NIST, ISO, NAIC); any quantified outcomes would be confirmed and substantiated in a formal proposal. Prepared independently by MCG/Kelly; not affiliated with, endorsed by, or originating from BCBSIL or HCSC.
Motion Consulting Group · a Kelly Services company Prepared for Blue Cross and Blue Shield of Illinois Confidential · Initial draft · June 1, 2026