Intro

Digital transformation has become one of those phrases that shows up in every boardroom, RFP, and strategy deck. But for many insurers, the reality is far from inspiring. Projects drag on, costs spiral, and the results rarely match the promise.
Part of the problem is how they start. Too often, transformation efforts are launched without a clear purpose, just the vague idea that being more “digital” is the goal. But without knowing what you’re trying to achieve, it’s nearly impossible to build a strategy that works.
Digital transformation isn’t about adopting the latest tech, but about rethinking how your business creates value. Without a clear goal and a strategy built to support it, transformation efforts tend to create more complexity, not more impact.
That’s why the story of Blue Cross Blue Shield of Michigan stands out. In one of the most highly regulated industries, they’ve done what many only aim for – they transformed. Not with hype or oversized teams, but with focus, discipline, and a sharp understanding of what needed to change – and why.
This article unpacks how they did it, and what insurance leaders can take away from their journey.

Who is BCBS Michigan – and why should insurers pay attention?

Blue Cross Blue Shield of Michigan is one of the largest nonprofit health insurers in the US. Headquartered in Detroit, they bring in $36 billion in annual revenue, have around 10,000 employees, and cover about 5 million people. While their name says Michigan, their reach is national. Most of their members live outside the state but work for Michigan-based companies.
That’s possible because they’re part of the broader Blue system, i.e., 34 independent Blue Cross Blue Shield companies that collectively serve over 115 million Americans. It’s a bit like a health insurance co-op – each one runs its own business but honors the others’ coverage. So if you’re a BCBS Michigan member and get sick in California, you’re still covered.
They’re also structured differently from many in the industry. BCBS Michigan is a mutual company, which means no outside investors and no pressure to deliver big profits. They typically operate on a razor-thin margin, about 1 to 1.5%, and reinvest what’s left into improving how they serve members.
Over the last decade, they’ve been through a quiet transformation. They used to focus mostly on employer-sponsored plans. Now, about a third of their business comes from government programs. That shift forced them to rethink how they operate – not just as an insurer, but as a broader health service platform.

Inside a transformation that didn’t fail

BCBS Michigan’s shift to becoming a full-scale digital health platform wasn’t just about keeping up with the times, it was about survival in one of the most complex, regulated industries in the U.S. And they started early.
Back in 2007, long before AI was trending, BCBSM began building artificial intelligence into their operations. Today, they’ve got more than two dozen AI applications running across the business. These tools help them do things like analyze member health risks, say, spotting early signs of diabetes or understanding complications in high-risk pregnancies. AI helps surface insights fast, so care teams and customer service reps can support members with real information, not generic scripts.
That’s the foundation. But the real leap came with generative AI.
For BCBSM, GenAI wasn’t just another tech buzzword. It became a strategic turning point. With just a dozen internal experts and a key partner, Forum Systems, they built a generative AI engine that sits on top of their data ecosystem. And because they already had the infrastructure – what they call their digital “chassis” – they could move fast.
The biggest challenge was the data itself.
In healthcare, around 90% of the most valuable data is unstructured. It’s voice calls, notes, emails, scanned documents, the stuff that doesn’t live neatly in spreadsheets or databases. Historically, that kind of data was too messy and expensive to use. But GenAI changed that.
Now, BCBSM uses GenAI to create “virtual data schemas” which essentially are digital frameworks that let them treat unstructured data like structured data. That means they can search it, analyze it, and act on it, without the heavy lift of cleaning or manually organizing everything first. Even voice data from customer service calls can be mined for insights.
And they’re not just experimenting, they’re deploying. In the last six months alone, they rolled out three real-world, AI-powered products:
1. Contracts
The first major GenAI use case was contract management. BCBSM has tens of thousands of contracts across the organization, covering everything from IT to clinical services. These contracts are long, detailed, and full of nuanced terms, which is a perfect territory for AI.
They fed the contracts into the system, built out a taxonomy, and trained a large language model (LLM) to understand and navigate it. Now, teams can ask natural-language questions like “Which of our contracts conflict with the new regulatory rule?” or “Where are we paying for the same service twice?”
It took a year to get the model working smoothly. Two years in, they’ve saved over $10 million by eliminating overlap and improving terms.
2. Security and trust
In healthcare, trust isn’t optional, it’s everything. So BCBSM built GenAI with security at the core. They call their environment Secure GPT, and it’s designed for high-risk, high-regulation industries like theirs.
It runs on three strict rules:

• Guardrails. These keep conversations on track, ensuring AI doesn’t go rogue or generate inappropriate content, which is especially critical in a clinical or customer service context.
• Data protection. No personal health data touches public LLMs. Everything sensitive is obfuscated before it moves anywhere.
• Observability and audibility. Every AI interaction is logged and traceable. They can reconstruct who accessed what, when, and why, with full forensic capability.

BCBSM now includes this framework in their RFPs, because customers ask the hard questions: “How do you govern AI? How do you handle bias? Who’s in control?” And they can answer with confidence.
3. Benefits at the point of care
One of the toughest parts of healthcare is figuring out what’s covered, what’s not, and how to actually get the care you need. BCBSM’s third GenAI use case tackles that head-on.
They’ve built an AI solution that helps members access benefit information in plain language, in real time, at the point of care. It can guide someone through their coverage, recommend services, and flag additional benefits they might not know about. All tailored to their specific situation.
It’s not just faster, it’s also smarter. And it’s changing how people engage with their healthcare, one interaction at a time.

From the outside, this might look like a tech upgrade… but it’s not

It’s a mindset shift. BCBS Michigan’s story isn’t just about AI or cloud infrastructure. It’s about rethinking what it means to serve people in a system that’s historically been anything but personal.
They used to be measured by how fast they processed claims or how many calls they answered. Now? It’s about how well they understand each member as a whole person, and how quickly they can act on that understanding.
This is where the industry is headed. If you’re in healthcare, or any complex, regulated field, the takeaway isn’t “go build an AI model.” It’s this – start by seeing your organization as an information platform. One that’s flexible, connected, and focused on real human needs.

How to do it right? What insurers can learn from BCBS Michigan

If you’re in insurance and trying to figure out how to actually do digital transformation (beyond the buzzwords), here’s the playbook. BCBS Michigan didn’t just implement new tools; they rethought how the business works. And they did it in a way that’s replicable.
Here’s how they pulled it off, and how you can, too:

1. Rethink your business model, you’re in the information business now

BCBSM understood something that many still miss – healthcare is no longer about pushing claims through the pipe. It’s about understanding people, deeply, and at scale.
That means collecting and connecting data across the full spectrum of a person’s life, from behavioral health and employment to housing and social support. The job of an insurer now is to use that data to support better outcomes, not just process transactions.
If your systems still treat members like policies instead of people, that’s the first place to start.

2. Lay the groundwork before you chase the trend

Before they rolled out any GenAI, BCBSM had already spent years building their digital foundation. They created what they call a “chassis”, a flexible data and application ecosystem that could support rapid experimentation and deployment.
The mistake many organizations make is skipping straight to shiny tools without the architecture to support them. Don’t start with GenAI. Start with infrastructure, i.e., clean data, integrated systems, and clear governance. Without that, your AI initiative will hit a wall before it ever gets off the ground.

3. Start small, stay sharp and choose the right partner

BCBSM didn’t launch this with a 200-person innovation team. They did it with 12 people. Experts, working in tight alignment with a single, capable partner in Forum Systems.
This means you don’t need scale to get started. You need clarity, discipline, and a partner who understands both your business and the tech. Big teams can bog down progress while focused teams move fast.

4. Use GenAI to unlock the messy stuff

Unstructured data is where GenAI really earns its keep. And in healthcare, that’s most of what matters, i.e., voice logs, contract PDFs, doctor’s notes, RFPs.
Digitizing all that data manually is way too slow and expensive. But BCBSM used GenAI to create virtual taxonomies – digital maps that make unstructured data searchable and usable, without restructuring it all first.
If your organization is sitting on piles of documents, transcripts, and call data, this is your opportunity. GenAI can turn that liability into a live asset.

5. Build trust into the system, not around it

Trust isn’t a nice-to-have, it’s the barrier between success and disaster. Especially in healthcare, where a single data breach can cost millions and break member confidence overnight.
BCBSM addressed this head-on by building Secure GPT, an internal GenAI platform with strict controls:

• Guardrails to keep responses appropriate
• Role-based access and purpose-specific use
• Full auditability, down to the user and session level

You can’t outsource trust. You have to architect it from day one.

6. Focus on business value, not just “innovation”

One GenAI app that scans contracts has already saved them $10 million by eliminating redundant agreements. Another helps members understand and use their benefits in real time. A third supports compliance and security at a scale that manual processes couldn’t match.
Whatever your AI strategy is, make sure it ties directly to outcomes like faster service, better care, lower cost. If you can’t measure it, don’t build it.

7. Train your people to use AI, not avoid it

No AI initiative works without buy-in. BCBSM made adoption a priority, not an afterthought. Within a week, they trained 170 employees to use their internal Secure GPT system.
It’s a powerful reminder that AI doesn’t replace people, it extends them. But only if they know how to use it; training isn’t optional, it’s a core part of the transformation.

Digital transformation checklist for insurers

Use this checklist to evaluate your readiness or guide your next steps:

Have you shifted from a transaction-based model to one that’s data- and service-driven?
→ Are you actively using data to understand members as people, not just policies?

Do you have a digital architecture that supports AI i.e., clean data, integrated systems, and scalable workflows?
→ Have you invested in building your “digital chassis” before layering in GenAI?

Are you starting with a small, dedicated team and clear scope?
→ Do you have the right partner who understands your business, not just the tech?

Are you applying AI to contracts, call transcripts, or clinical notes?
→ Have you explored virtual taxonomies or schema strategies for hard-to-reach data?

Do you have strong AI governance in place – audit trails, role-based access, and privacy controls?
→ Are your teams clear on what AI can and shouldn’t do?

Can you tie your AI or digital initiatives to measurable outcomes?
→ Are you solving real pain points like cost savings, compliance, or member experience?

Are employees equipped to use AI tools effectively, not just avoid them?
→ Do you have a training plan in place to scale adoption?

Transformation isn’t about size, it’s about focus

You don’t need to be the biggest insurer in the country to pull this off. But you do need to get the fundamentals right which is the business model, architecture, trust, and people. BCBS Michigan didn’t just modernize, they changed how they think about what a health insurer is.
You can do the same.
If you’re looking for a partner on that journey, Clurgo brings the technical depth and strategic focus to help make it happen. With experience guiding highly regulated industries through complex digital transformations, Clurgo can help you build the right foundation, and scale the right innovations. Like BCBSM, you don’t need a giant team – you need the right one.