We can now predict an employee's risk of developing type 2 diabetes, cardiovascular disease, or metabolic syndrome months or years before symptoms appear. The science works. The models work. What doesn't work is the gap between what we can predict and what organisations are prepared to do about it.
What You Need to Know
- AI-driven predictive health models can identify disease risk from biomarker patterns, genomic data, and lifestyle signals with clinically validated accuracy
- Most employers don't have the infrastructure, policies, or cultural readiness to act on predictive health insights
- The ethical questions are real: who owns the prediction? What's the employer's obligation? How do you avoid discrimination?
- Organisations that get ahead of these questions now will have a significant advantage in workforce health and retention within five years
What Predictive Health Can Actually Do
This isn't speculative anymore. Predictive health models are in clinical use. They're analysing combinations of biomarkers, genetic risk scores, lifestyle data, and environmental factors to flag risk before the body shows symptoms.
87%
accuracy achieved by machine learning models in predicting type 2 diabetes onset 5+ years before diagnosis using routine blood biomarkers
Source: Ravaut et al., Nature Scientific Reports, 2021
I've spent years in this space, first building precision health platforms at Edison, now working on corporate health strategy at UniMed. The capability curve has been steep. Five years ago, predictive health was an academic exercise. Today, it's a clinical tool.
AI models can identify patterns in routine blood work that a human clinician wouldn't catch, subtle shifts in metabolic markers that indicate the body is moving toward dysfunction years before a diagnosis. Combined with genetic risk scores and lifestyle data, these models create a health trajectory, not just a snapshot of today but a projection of where someone's heading.
The technology is ready. The question is whether workplaces are.
The Readiness Gap
When I talk to employers about predictive health, the conversation follows a predictable arc. Enthusiasm first: "Imagine if we could identify employees at risk and intervene early." Then the questions start.
"Who pays for the testing?" Comprehensive health assessments with biomarkers aren't covered by standard employment health checks. Someone needs to fund the data collection that feeds the models.
"What happens when we know?" If a predictive model identifies an employee as high-risk for a condition, what's the employer's obligation? Providing information is one thing. Providing a pathway to clinical intervention is another. Most employers don't have that infrastructure.
"How do we avoid discrimination?" Predictive health data in the wrong hands, or the wrong framework, creates serious risks. An employer who knows an employee's health trajectory could, intentionally or not, factor that into career decisions. The legal and ethical guardrails need to be built before the data starts flowing.
"What if employees don't want to know?" Not everyone wants to learn they're at elevated risk for a serious condition. Autonomy matters. A predictive health programme needs robust opt-in frameworks that are genuinely voluntary, not "voluntary but your manager really thinks you should."
These aren't reasons to avoid predictive health. They're reasons to get the framework right before deploying it.
What Ready Looks Like
The organisations that will benefit most from predictive health are the ones building readiness now. That means four things.
A clinical partnership, not just a wellness vendor. Predictive health insights need clinical pathways. When the model flags a risk, the employee needs access to a clinician who can interpret it, provide context, and recommend action. A wellness app notification saying "you're at elevated risk" without clinical support is irresponsible.
Privacy architecture that's actually robust. Health data is the most sensitive data an employer can hold. Predictive health data, which tells you what someone might develop in the future, is even more sensitive. The privacy framework needs to be designed for this specific context, with clear boundaries on who can see what, strong anonymisation for aggregate analysis, and genuine consent mechanisms.
A culture of health investment, not health surveillance. Employees need to trust that predictive health is being used to help them, not to assess their value. That trust is built over time, through consistent action that prioritises employee outcomes. Organisations with a history of genuine wellness investment have a significant advantage here.
62%
of employees said they would participate in employer-sponsored predictive health screening if they trusted data would not affect employment decisions
Source: Deloitte Health & Wellbeing at Work Survey, 2024
Leadership that models participation. When the CEO does the health assessment and talks openly about what they learned, it changes the dynamic. Predictive health can't be something the organisation does to employees. It has to be something the organisation does with them.
The Five-Year Window
I believe we're in a window right now, roughly 2025 to 2030, where predictive health will move from niche to mainstream in corporate settings. The science is validated. The AI models are improving rapidly. The cost of testing is dropping.
The employers who build readiness now will have healthier workforces, lower healthcare costs, and a genuine competitive advantage in attracting talent that values long-term wellbeing over short-term perks.
The employers who wait will eventually adopt predictive health too, but they'll do it reactively, in response to competitors or regulation, without the cultural foundation that makes it effective.
Sound familiar? It should. That's exactly the pattern we've seen in every major workplace health shift, from occupational safety to mental health support. The organisations that lead are the ones that build the framework before the mandate arrives.
The future of workplace health isn't a gym subsidy and an EAP. It's knowing that an employee is drifting toward a serious condition and having the infrastructure to help them change course.
Jay Harrison
Health Technology Advisory
Don't wait for the mandate. Build the readiness now.