May 17, 2024Phil Newton, Associate at law firm, Pinsent Masons, and Karl Simons OBE, Chief Futurist at technology firm, FYLD, delve into the potential impact innovative technologies can have on H&S management and legislative compliance.
We continue to see an increasing number of progressive organisations adopting artificial intelligence (AI) to bring efficiencies in safety, environmental sustainability, quality, and productivity.
Along with its impact, we will also look at the opportunities AI can bring to this space, including a much-needed proactive approach to H&S governance, as well as the extra responsibilities which may result. These include new and emergent risks and liabilities associated with the use of new technologies and the need to upskill people in their failure modes.
Traditional human-led, lead-and-lag safety indicators have been used by the H&S community for decades, but it is increasingly not enough to take organisations to the next level of incident and injury reduction. ‘Digital Safety Performance Indicators’ will become the norm for many H&S professionals working with the latest cutting-edge technology.
Phil Newton, Associate at Pinsent Masons
We are on the brink of a transitional phase where leading organisations are using a mix of traditional and digital safety performance indicators which combined are already showing companies achieving better reductions in incidents and injuries as a result.
As data-driven technologies become the norm, it is becoming standard to both identify the critical metrics for H&S performance and understand how analytics and AI are driving a new era of effective measurement. The evidence demonstrates that this approach significantly improves legislative compliance and an organisation’s ability to meet and exceed its targets.
Behavioural safety has been a major discussion topic for many years within the H&S community, as practitioners are acutely aware that affecting behaviour positively impacts on compliance with company policies and safety management systems.
It also ultimately reduces the likelihood of incidents and injuries occurring. AI solutions are already showing significant benefits in affecting behaviour to bring legislative compliance. The use of video to undertake ‘point of work’ risk assessment forces fieldworkers out of the van or site hut to physically be at the point of work, and then have to talk through and film the works area. This ensures far richer content is captured which is relevant to the works being undertaken.
In addition, as all field teams are making visible their start of works assessment, this subconsciously ensures they know they must complete the assessment as their line manager is expecting to receive the real-time digital notifications via their own device (i.e. a push notification). Naturally, this then drives up compliance levels through management having confidence all teams have completed a suitable and sufficient risk assessment prior to the commencement of work.
Competency has always been attributed to a person’s knowledge, skills, and experience. This has a natural human limiting factor.
The new wave of digital solutions is aiding workforce development because new technologies bring collective knowledge to the workforce, which is no longer reliant on human-to-human training, and they are able to coach workers to improve their skill set.
The advancements in AI have allowed humans to extend their capabilities and accuracy, and at times with a reduced workforce or without exceptional talent, the use of AI can still produce superior results. This has made the notion of high-quality accessible data a realistic enhancement to a wide range of people and companies.
Some organisations are recruiting AI-specific professionals such as a ‘Chief AI Officer’ or an AI team. The competency and training of H&S professionals is therefore a key consideration for organisations as this new technology is adopted.
There is the potential for a legal compliance risk to arise when organisations don’t act on the safety data that is being generated. As organisations incorporate AI systems into their business, it is imperative they also develop the competency of their people to analyse the volume and complexity of data these systems generate and act to minimise any safety risks identified.
The classic legal argument is that organisations are not required to ‘man-mark’ or watch over the shoulders of every employee.
The level of supervision of course depends on the age, skills, knowledge and experience of the worker. Technology can be a useful tool in helping organisations take reasonably practicable steps to supervise employees and complement more traditional methods that require the physical presence of those undertaking supervision or monitoring roles.
Technology has enabled remote visibility of workers. Although for an experienced H&S practitioner, there will always be the need to balance the benefit of adopting technology to increase visibility aimed at enhancing safety management controls against the increased risk of having gathered higher volumes of data.
This can lead to missed opportunities to intervene or capture data that demonstrates action should have been taken.
In a recent corporate manslaughter trial, there was CCTV of the incident itself (which was compelling evidence of the breach) but the company could not provide a valid reason for having installed CCTV other than to make sure workers were doing their job. In effect, the organisation had inadvertently assumed legal risk without properly thinking about the rationale and what steps they should take as a consequence of having that level of surveillance installed.
Karl Simons OBE
AI and digital technologies can connect workforces through effective communication and collaboration. AI can effectively collaborate with human workers, bringing information to their attention.
Modern technologies can expand the range of communication from manager to team(s) where information can be shared and broadcast immediately at the point of need.
The use of techniques, such as Natural Language Processing and Imagery Analysis, can enable better compliance with legislation through translation tools that connect both worker and line manager who perhaps speak different languages. No longer will we see signs written in multiple languages – it will be the digital signage of the future that recognises and communicates with a worker in their own tongue.
Under The Health and Safety at Work Act employers must inform and consult with their employees on measures to be introduced that may affect their health and safety.
These measures have traditionally been communicated to employees via policies and procedures but AI, when used correctly, can be a useful tool in enabling far greater communication of information to aid the worker in assessing risk inherent with their role.
This will also help organisations demonstrate to an enforcing authority that safety measures have been effectively communicated to the workforce, moving beyond the ‘evidence’ of induction record tick boxes.
The adoption of AI also raises an interesting question with regards to the legal principle of foreseeability – if an AI system can predict incidents, but action isn’t taken on the information behind that prediction, organisations could leave themselves exposed to the argument that they have failed to prevent exposure to a reasonably foreseeable risk of harm.
This is why smart AI solution providers right now are using a blend between AI-driven responses that are then augmented by human oversight and final sign off by the employer on what is produced by the system.
The future workplace where computers are solely responsible for development and final sanctioning, removing the element of human oversight, could introduce unforeseen risk to an organisation without any fail-safe systems.
Furthermore, there is a risk that AI will identify previously unforeseeable risks and redirect time and resources to their management and away from the key safety risks facing the business. The law requires employers to take all reasonably practicable steps to eliminate or mitigate foreseeable risks to health and safety, whether or not they have been identified using AI tools.
One of the immediate impacts of a H&S incident is that it kickstarts a forensic analysis by an organisation (and potentially the regulator) into the adequacy of the safety management system in place at the time.
AI solutions can assist organisations in these circumstances by demonstrating to a regulator the systemic approach taken to legal compliance. A systemic failure to address risks to H&S is a factor that could put an organisation in the ‘high culpability’ bracket under the relevant sentencing guidelines so adopting an AI solution will help to mitigate that legal risk.
There are also potential evidential benefits to organisations who adopt AI solutions from the inherent site and task specific nature of the technology. The evidence that can be extracted from AI solutions will help organisations demonstrate their understanding and assessment of risk in the workplace or in its monitoring of the workforce. At the same time, there are potential risks to organisations following an incident where AI is involved.
The Police and regulator have always had the power to request pre-existing material, whether written or computer-held, as part of their investigation. AI solutions by their very nature generate data which is stored on servers and will be a legitimate source of evidential material. This data will not be protected from disclosure by legal privilege as it is material that came into existence prior to the incident.
There is also the risk that a legal duty holder will fail to act on the enhanced information that AI solutions can generate. The general duties under the Health and Safety at Work Act are positive duties in that they require employers to act.
Organisations also need to have a clear understanding of how safety data generated by AI solutions impacts on their understanding of the likelihood of harm occurring – another factor relevant to H&S sentencing.
Organisations are continuing to embrace digital solutions and the use of AI technology is becoming normalised as we seek to make our workplaces safer. There remain compelling arguments both for and against its use.
Misuse of technology has and will continue to create increased risk to organisations should proper safeguards not be put in place. However, use of information derived by AI can and is aiding humans in evidencing legislative compliance and achieving incident, injury and illness prevention.
Quote:
“As organisations incorporate AI systems into their business, it is imperative they also develop the competency of their people to analyse the volume and complexity of data these systems generate and act to minimise any safety risks identified”
