Eliminating barriers in reporting: how conversational AI can unlock safety engagement

Walk into almost any organisation with a mature health and safety programme and you will find the same story repeated: robust systems, clear processes, well-designed forms, and dashboards that track everything from near misses to nonconformities. On paper, the infrastructure for reporting is there. And yet, when you look at the actual volume of reports coming through, something doesn’t add up. 

This isn’t a systems problem. It’s a human one. And if we want to close the gap between the safety culture we say we have and the one that actually exists on the ground, we need to understand why people don’t report and remove those barriers, rather than just hitting the same note telling people to use the system.

Barriers to reporting
In my conversations with safety leaders across a wide range of industries, three barriers to reporting continually came up. First, there’s been a lack of meaningful training. Many employees genuinely don’t know they’re expected to report hazards or why it is so important. Workers may understand that reporting matters to the business – ticking the boxes of compliance, audits and insurance – but that framing does little to motivate someone standing next to a frayed cable or a blocked fire exit. Reporting needs to be understood as something that protects them, not just the organisation. When people see a direct line between speaking up and their own wellbeing (or that of their colleagues), participation changes.

The second barrier is the forms that feel like a chore. Even well-intentioned reporting systems often ask for too much information, such as category codes and text boxes that demand a level of detail most people don’t have time to provide in the middle of a busy shift. For someone standing on a factory floor or a construction site, a five-minute form can feel like a genuine obstacle, especially if they’ve already clocked what needs to happen and just want someone to fix it. The longer and more complex the form, the more likely it is that the report simply doesn’t happen.

Lastly, there is the fear of getting it wrong, or getting in trouble. This is perhaps the most persistent barrier of all. Workers worry about being blamed, about implicating a colleague, about looking like they don’t know what they’re doing, or about challenging a supervisor’s judgement. 

I had a very interesting conversation with Graham Goulden on our webinar about this. Graham is a former chief inspector with Police Scotland who now works as a consultant, training organisations on bystander intervention. He argues that hierarchy should be left at the door when it comes to health and safety, but organisations must build a clear people-centric approach to get there. 

Most reporting systems were built with the organisation in mind: what data does the business need to capture for compliance, audits and trend analysis? That’s understandable, but it puts the burden of translation on the worker: they are the ones who have to take a lived experience and turn it into structured, categorised data that fits neatly into a form built for a compliance team, not for them.

The potential of conversational AI
A people-centric approach flips that relationship. Instead of asking workers to adapt to the system, the system adapts to how people naturally communicate. It’s a simple idea, but there are significant implications for how safety technology should be designed. That’s where conversational AI has real potential.

Most people find it easier to speak than to write. That’s not a controversial claim; it’s borne out in how we communicate every day, and it’s especially true under time pressure or in a physically demanding environment, where typing on a phone screen with gloved hands, or standing in noise and dust, is genuinely impractical.

Conversational AI allows workers to report a hazard by simply speaking into a mobile app, the way they might describe the issue to a colleague. No dropdown menus. No mandatory fields. No wondering which category a leaking hydraulic hose is supposed to fall under. The worker just talks, and the system does the work of structuring that information behind the scenes. 

This doesn’t mean sacrificing the detail organisations need. If the AI recognises that a report is missing key information, such as the location, time, or whether anyone was injured, it can simply ask. That distinction matters enormously when one of the core barriers to reporting is intimidation.

A conversation feels lower stakes than a formal document, eliminating the sense that a single poorly chosen word will be scrutinised later. Speaking naturally, in your own words, and having the technology handle the structuring removes a layer of psychological friction that a static form simply can’t.

Bridge generational differences
Another interesting aspect of conversational AI in this context is how well it can bridge generational differences in a workforce, something safety technology often struggles with.

For younger workers who have grown up with voice assistants, chat interfaces and conversational technology as a normal part of daily life, speaking to an app to report a hazard requires no adjustment at all. It’s simply how they already expect to interact with digital tools. For older workers, or those less confident with technology generally, the benefit comes from a different direction. Rather than needing to navigate an unfamiliar interface with multiple screens and menus, they can simply describe what they’ve seen, in their own words, the same way they’d describe it to a supervisor face-to-face. 

Again, this is where the organisation should emphasise its people-centric culture. It may seem contradictory to use an AI tool and still reinforce that we put people first, but that’s what we discussed in one of our recent webinars with Richard Bate, president of IOSH: AI is there to support human work, not to replace it, and that needs to be clear to leaders and team members. 

If we resist new technologies completely, we’re getting left behind. But if we keep learning and using what’s new out there to our advantage, everybody wins.

Importance of feedback
There is one other solution that resonates with people from different generations: feedback. 

It doesn’t matter what the age or background of the worker is: if they feel heard, they’re going to speak up again. If they report a nonconformity and get positive feedback or even have the opportunity to follow up on the incident’s investigation, they’ll become more confident about repeating the process. And even better: they might become ambassadors of safety and help engage their colleagues, building a genuine culture around it.

Ultimately, safety reporting was never meant to be a bureaucratic exercise. At its core, it’s a form of communication: one person telling another, “I saw something that could hurt someone.” Somewhere along the way, in the effort to standardise and structure that communication for organisational purposes, we made it harder for people to do the one thing we actually wanted from them.

By allowing people to report the way they already communicate, by describing what they saw in their own words, organisations can lower the barriers that have quietly suppressed reporting for years. Technology doesn’t replace a strong safety culture, but it does remove the friction that so often stands between that culture and the everyday behaviours needed to sustain it.

If we want people to speak up, we need to make it as easy as simply speaking.

ecoPortal is a health and safety management software that helps organisations streamline incident reporting, risk tracking, audits and compliance. Julian Taylor is senior enterprise account executive at ecoPortal, and hosts monthly online sessions focused on modern health, safety, wellbeing and leadership challenges as a part of the organisation’s ‘Safety Voices’ webinar series.

For more information, visit: 
ecoportal.com