State of AI Risk by Role — March 2026

Every month, thousands of professionals complete a Runway assessment. That data paints a picture of AI career exposure that no analyst report can match — because it comes from the people actually doing the work, not from executives guessing about it.

Here's what we're seeing in March 2026.

The headline numbers

Across all assessments completed in the last 90 days:

• Average automation risk: 47/100 — up 3 points from December 2025
• Average defensive strength: 41/100 — roughly flat
• Near-future risk gap: +8 points — the gap between today's risk and 24-month projected risk is widening

That last number is the one that matters. The gap is accelerating.

Roles with the fastest-rising risk

These roles saw the largest increase in automation risk scores over the past quarter:

1. Junior data analysts — risk up 11 points. The combination of structured data tasks and improving LLM reasoning is compressing this role fast.
2. Marketing coordinators — risk up 9 points. Content generation, campaign reporting, and social scheduling are all in the blast zone.
3. Paralegal / legal assistants — risk up 7 points. Document review and contract analysis tools crossed a quality threshold in Q1.
4. Customer support managers — risk up 6 points. Not the frontline agents (already affected) — the managers whose value was routing and quality control.

Roles more protected than expected

Not everything is accelerating. Some roles are proving more resilient than the narrative suggests:

• Senior sales executives — risk down 2 points. Relationship depth and negotiation complexity remain difficult to automate.
• Specialist nurses — risk stable. Clinical judgment + physical presence + regulatory requirements create a triple moat.
• Construction project managers — risk down 3 points. The gap between digital planning and physical execution is a natural defence.

The defensive strength problem

Here's the uncomfortable finding: 62% of users with risk scores above 55 have defensive strength below 40.

That means the majority of people most at risk have the least protection. Not because they can't build defences — but because they haven't started yet.

The most common defensive gaps we see:

1. No AI fluency — 34% of high-risk users report using zero AI tools in their workflow
2. No domain specialisation signal — generalists with broad-but-shallow skills
3. No transition readiness — financial runway under 3 months, no support network

What this means for you

If you haven't taken an assessment recently, these numbers might feel abstract. They shouldn't.

The 24-month risk gap is the number to watch. It tells you how much worse things get if you do nothing. For most roles, that number is between 5 and 15 points. For some, it's 20+.

The good news: defensive strength is entirely within your control. Every point you add buys time. And time is the one resource you can't get back.

---

This analysis is based on anonymised, aggregated data from Runway assessments. No individual data is exposed. Want to see where your role sits? Take the assessment.