Home » Market Insights » Workplace Safety & Risk » EHS Data Reveals Increase in Work Injuries | Benchmark Gensuite

Workplace injury frequency and severity are climbing two to three times faster than last year — and the data points to why

Benchmark Gensuite’s 2026 EHS Benchmarking Report shows injury rates rising sharply while 90% of incidents go underreported. The headline numbers tell part of the story — what’s actually driving them is a workload problem, not a commitment problem, and the report points to where AI has and hasn’t closed the gap.

published

February 18, 2026

Originally covered by

TOPIC

Workplace Safety & Risk

Original research

R. Mukund
Founder & Chief Executive Officer, Benchmark Gensuite

As covered in

ISHN │ OHS Canada │ Work Safety 24/7 │ HCL Review │ Health & Safety International │ Facility Safety Management

━ Overview

ISHN, OHS Canada, Work Safety 24/7, Human Capital Leadership Review, Health & Safety International, and Facility Safety Management all picked up the headline findings from Benchmark Gensuite’s 2026 EHS Benchmarking Report this week — injury frequency and severity climbing sharply, underreporting at critical levels, and EHS teams absorbing responsibilities that have expanded well beyond traditional safety duties. This page goes further into what more than 260 EHS professionals actually told us: not just what’s rising, but why, and what’s really behind the gap between AI adoption and AI impact.

The numbers are the headline. The workload math underneath them is the story EHS leaders need to understand.

━ Benchmark Gensuite Perspective

The injury spike isn't a safety failure — it's a workload failure

It would be easy to read a two- to threefold increase in reported injury frequency and severity as a sign that safety programs are breaking down. The data doesn’t support that read. What it shows instead is a workforce being asked to do more — sustainability reporting, emergency preparedness, total worker health, supply chain compliance — without a corresponding increase in the people or hours available to do it. Safety commitment hasn’t declined. Capacity has been stretched past the point where manual, reactive processes can keep up.

Increased demand and workforce shortages now top the list of factors leaders point to when explaining rising incidents, ahead of time constraints and training gaps. That ordering matters. It says the primary driver isn’t a skills problem or a tooling problem in isolation — it’s a bandwidth problem. Every hour spent on cross-functional ESG coordination or emergency plan updates is an hour not spent on the field walkthroughs, hazard observations, and proactive risk work that actually prevents incidents.

The underreporting figure is the clearest evidence of that strain. When 90% of incidents, hazards, or near misses go undocumented, the problem usually isn’t a lack of policy — it’s that reporting competes for the same limited time as everything else on an EHS professional’s plate. And underreporting isn’t a neutral gap. It removes the early warning signals that let teams intervene before a near miss becomes an injury, which is very likely part of why frequency and severity are both moving in the wrong direction at once.

“EHS leaders are battling rising injury rates as demand skyrockets and the workforce dwindles. With new responsibilities further complicating roles, teams are strapped for time when it comes to proactive risk mitigation.”

— R. Mukund, Founder & CEO, Benchmark Gensuite

This is also where the AI adoption numbers need a second look. 92% of EHS leaders report using generative AI and 86% report using agentic AI — adoption most organizations would call a success story. But the report shows that use is still concentrated in administrative tasks: summarizing incident reports, supporting audit documentation. Valuable time savings, but not yet applied to the operational workflows — hazard triage, predictive risk flagging, frontline decision support — where injury prevention actually happens. The adoption curve has moved faster than the operational integration curve, and that gap is exactly where the injury data is still climbing.

45%

Report increased injury frequency, up from 18% last year

90%

Say incidents, hazards, or near misses go underreported

92%

Report using generative AI in day-to-day work

None of this argues for doing less measurement or fewer safety programs. It argues for closing the gap between where AI has landed so far and where the actual risk decisions get made — which is the throughline the rest of this data points to.

━ Key themes

Three things to understand about the 2026 injury data

1

Rising injury rates are a capacity signal, not a culture signal

Increased demand (44%) and workforce shortages (42%) outrank time constraints and training gaps as the top drivers leaders cite. The data points to teams that are stretched thin, not teams that have stopped caring about safety outcomes.
2

Underreporting is a visibility problem with a compounding effect

90% underreporting means the data organizations act on is already incomplete before analysis even starts. 39% of leaders say they're missing early signals — near misses, worker feedback — that would otherwise flag risk before it becomes an injury.
3

AI adoption has outpaced AI's operational impact

Near-universal generative AI use hasn't yet translated into fewer injuries, largely because current use is concentrated in documentation and summarization rather than the frontline workflows where risk decisions actually happen. The report frames this as the next frontier, not a failure of the technology.

━ What this means for EHS leaders

Operational implications

EHS leaders managing stretched teams should treat the demand-vs-headcount gap as the root cause to address, not the injury numbers themselves. Reallocating administrative overhead — documentation, report summarization, compliance tracking — away from manual effort is what actually frees capacity for the field work that prevents incidents.
Compliance and reporting managers face a data quality problem before they face an analysis problem. With up to a quarter of employees not reporting at all by some estimates, the priority is lowering the friction and time cost of reporting itself, not just building better dashboards on top of incomplete data.
Sustainability and ESG leads absorbing EHS-adjacent duties are a documented factor here — 46% of EHS leaders report added oversight tied to environmental and governance work, more than a quarter without added resources. That convergence needs to be planned for explicitly, not treated as incidental overflow onto existing teams.
Senior leadership and executive sponsors should read the AI adoption numbers as a call to invest one tier deeper — beyond drafting and summarization tools already in wide use, into the predictive and workflow-level AI that acts on risk signals before they become incidents. That’s where the injury curve is most likely to actually bend.

━ Next steps

What organizations should do now

The instinct when injury numbers rise is to add more oversight — more forms, more reviews, more manual checkpoints. This data suggests that’s the wrong lever. The teams reporting the sharpest increases are already stretched past what manual oversight can absorb. The more productive move is redesigning where time goes, not adding more demands on the same hours.

Start by mapping where your team’s AI use currently sits — if it’s concentrated in drafting and documentation like the benchmark average, that’s the signal to look at operational and predictive use cases next. Audit your reporting process specifically for friction: if submitting a near miss takes meaningfully longer than the value an employee sees in doing it, underreporting will persist regardless of policy. And treat sustainability and governance responsibilities landing on EHS teams as a resourcing decision, not an assumption.

If you’re a current Benchmark Gensuite subscriber, the workflows to close this gap — automated triage, predictive alerts, streamlined reporting — are available in your existing platform. If you’re evaluating how your organization stacks up, the full 2026 EHS Benchmarking Report includes the complete methodology and findings referenced here, and is worth reviewing before setting next year’s safety priorities.

━ Featured executive

R. Mukund

Founder & Chief Executive Officer, Benchmark Gensuite , Benchmark Gensuite

The Benchmark Gensuite platform was built under Mukund's leadership from inception. Today it supports 480+ companies globally as subscribers of its cloud-based, best-practices-driven Environmental, Health, and Safety digital business transformation software — spanning risk and compliance, sustainability and disclosure reporting, frontline operations, quality, and product stewardship. The platform serves over 8 million workers across 30+ industry sectors and 150+ countries, supported by a global team across 10 offices, and headquartered in Mason, OH.

Media Coverage

Official press release

Read Benchmark Gensuite's original announcement of the 2026 EHS Benchmarking Report findings.

2026 EHS Benchmarking Report

Get the full findings and methodology behind the data referenced on this page.

About the 2026 EHS Benchmarking data

Why are workplace injury rates increasing in 2026?

Benchmark Gensuite’s 2026 EHS Benchmarking Report found that 45% of EHS leaders reported increased injury frequency, up from 18% the prior year, and 39% reported increased severity, up from 13%. Increased demand and workforce shortages were identified as the leading contributors, followed by time constraints and insufficient training.

What is causing incident underreporting in EHS?

90% of EHS leaders report that incidents, hazards, or near misses are going underreported, up from 79% the previous year. Nearly half estimate that up to a quarter of employees don’t report incidents at all, often because reporting itself competes with an already stretched workload.

How many EHS leaders are using AI in 2026?

92% of EHS leaders report using generative AI and 86% report using agentic AI in day-to-day work. Generative AI is used most for summarizing incident reports and supporting compliance documentation, while agentic AI is applied to predictive insights, automated alerts, and report triage.

What did Benchmark Gensuite's 2026 EHS Benchmarking Report find?

The report, based on responses from more than 260 EHS professionals, found rising injury frequency and severity, critical levels of incident underreporting, expanding sustainability and emergency preparedness responsibilities without added headcount, and accelerating AI adoption concentrated in administrative rather than operational workflows.

Why is EHS workload increasing?

EHS teams are absorbing sustainability reporting, emergency preparedness, and total worker health responsibilities on top of traditional safety duties, largely without corresponding increases in headcount. 46% of EHS leaders report additional oversight tied to environmental and governance expectations, and over a quarter say this growth has come with no added resources.

Is generative AI actually reducing workplace injuries?

Not yet at scale. The 2026 EHS Benchmarking Report found that most current AI use is concentrated in administrative tasks like documentation and summarization rather than the operational workflows where risk decisions get made. Benchmark Gensuite’s position is that AI needs to move into frontline risk workflows, not just back-office tasks, to move the injury data.

SEE IT IN ACTION

Ready to see Benchmark Gensuite in action?

See how enterprise EHS teams use our platform to manage risk, compliance, and safety outcomes at scale.