The safety and health field “is chomping at the bit” to adopt artificial intelligence, Tom Goodmanson, CEO of EcoOnline, a safety and sustainability software company, tells ISHN. “It’s ripe for adoption.”

That assessment as 2026 begins is echoed in survey data from the What Works Institute and Evotix, an safety and health software supplier. John Dony, CEO and co-founder of the institute, reports that most safety and health programs surveyed are piloting AI in limited scopes (42%) or exploring uses (33%). Less than one in ten (8%) are adopting AI on a broad scale, and about one in five (17%) have not started an AI journey.

Accelerating AI interest that has “reached the tipping point” is due to several factors, according to Goodmanson:

• Safety and health programs have matured beyond policing, reactionary practices, spreadsheets and paper records;
• More companies have matured, too, emphasizing a cultural value in keeping their people safe;
• The maturation of professionals and their companies is matched by maturing technology, resulting in increasing receptiveness to use AI technology for safety and health;
• More AI technology tools designed for safety and health applications are now available than ever before;
• Compliance, still the mainstay of safety and health work in many companies, has lagged in tech adoption and is “ripe for technological efficiency.”

Tips for safe implementation 

The appetite for AI is evident, according to an EcoOnline survey of 175 safety and health pros — 45% see clear potential for AI applications and 20% are open-minded about AI deployment, saying it depends on implementation. As many pros weigh the opportunity of AI, Goodmanson raises several issues he believes need to be addressed early on in AI journeys to ensure safe implementation:

• Pros need to be in lockstep with their company’s overall AI adoption plans. Don’t confine AI uses to a safety and health silo. Where is your company at on its AI journey? It’s critical to see and understand the bigger picture.
• How good is your data? Take a hard unbiased look. AI’s ability to analyze and predict is only as robust as the data set it searches. If your incident, auditing, observation and other reporting is thin or shallow, the chances increase for AI to make inaccurate predictions, risk assessments and image analyses. Are your metrics mostly lagging indicators or do they include leading indicators? Lagging measures alone will limit and a blend of leading/lagging measures will empower AI algorithms.
• How have you engaged your front-line people regarding AI adoption? Workers will know when AI gets something wrong, and when it genuinely adds value. Launching AI initiatives without workers understanding and providing input on applications can lead to wasted time and money.
• How do you talk to senior leaders about investing in AI for safety and health? These conversations should be approached with openness: Are you comfortable with the possibility of investing in AI instead of hiring more staff? And with assertiveness: Senior leaders need to understand there will be the need for the human in the loop; AI is not replacing human beings — Goodmanson is emphatic here. “AI enhances safety and health skills, it gives pros a leg up,” he says.  

Still, there might be a case, for example, where you have 50 people conducting audits and you would like to add another 50. AI using rebuilt protocols might enable those 50 people to double their output. The skills of the 50 workers already doing the work are enhanced rather than adding another 50 for auditing. As with other business disciplines — computer programming, administrative work, customer service, accounting and auditing, sales — AI in some cases can reduce the number of safety and health practitioners needed, but it will never replace the human in the loop.

Safety and health work, like customer service, nursing, front line supervision, maintenance and repair work and even executive management, will be augmented by AI by taking over routine tasks.

Questions to ask yourself

• Can you spot flashy AI? As products continue to roll in to the safety and health market with aggressive marketing, you need assurance AI is not giving a false sense of security. Have tools been tested to ensure there is not a gap between AI and reality? You must have confidence AI is not giving a recommendation that all’s safe when people are at risk. “Vendors need to be able to demonstrate the capabilities and accuracy of their AI in person and in real time, not just describe how good they are,” says Goodmanson. “Like Wendy’s said in the 1980s, you need to be shown the beef.”
• Are you ready to be a leader in the AI journey? To do so, you need to be informed by connections in your company where the organization is at in terms of AI implementation. Has a certain maturity been reached? Goodmanson says large companies are more likely to look at AI as a reality, while mid-size companies are more curious and not as far along with a comfort level for AI uses. Leaders are aware, savvy, curious, knowledgeable about tools on the market, and understand how AI can make the organization safer.
• Are you ready for a change in how safety and health is practiced? The nature of work is changing everywhere in the business world, and safety and health is no exception. Safety and health departments might retain the same size of staffing or possibly require fewer heads in the case where AI is broadly scaled up.

Again, this is a stage of implementation that only a few organizations have reached in 2026.

Are you prepared to coordinate and analyze more data points coming from more locations, or from more individuals using wearable AI for monitoring PPE use, lone worker safety, and personal health metrics? Have you built the trust needed for workers to be comfortable with “big brother” surveying over their shoulder and to be confident that health data privacy is assured? Are you ready to develop the expertise needed to write prompts, questions, for AI to answer regarding visual imagery, investigations, spotting patterns, proposing corrective actions and training based on recent incidents or audit findings?

To date there has not been widespread serious investment in AI for safety and health apps, but a gradual, consistent process will continue in 2026 to make AI a reality for more professionals and their companies.