We are witnessing the rise of Artificial Intelligence (AI), capable of processing vast amounts of data and learning from its interactions.
AI and Workplace Safety
AI's potential to enhance occupational health and safety is immense. It can analyse extensive data, identifying crucial patterns and trends that would take humans hundreds of hours to find. For instance, AI can scan an organisation’s safety reports, providing insights and highlighting trends related to near-miss observations and accidents. This capability was demonstrated in a study by Lund University, where AI matched the accuracy of skilled radiologists in identifying signs of cancer in breast screening tests, effectively substituting five months of a radiologist’s work.
AI-based monitoring tools can plug into organisational CCTV systems, identifying hazardous events and patterns, such as near-misses and poor manual handling practices. This advanced surveillance provides safety professionals with rich datasets, allowing them to identify hazard hotspots without the need to scan hundreds of hours of footage manually.
However, the use of AI in monitoring raises concerns regarding the handling of sensitive personal data. Transparency in collecting and using such data is crucial, and workers should have access to the information collected. AI should be used to improve arrangements or training and involve the workforce in solving problems, not to support a “blame culture”.
AI and Decision-Making
The evolution of AI into agentic AI, which combines large language models with functional software and hardware, allows AI to make and execute decisions. This development promises savings in human effort and time, handling tasks ranging from marketing campaign scheduling to copywriting and monitoring responses. However, the potential for AI to control physical systems has led to calls for greater regulation due to the risk of misuse or unintended consequences.
Chat-based AI like ChatGPT and Bing Chat are primarily used for text-based research and communication. They are trained on large volumes of publicly available web pages, synthesising results in various text formats. However, their reliability in providing risk management information is questionable, as they cannot adjust for local circumstances beyond their database and do not offer a potential margin of error for any information they serve up.
AI and Automation
Advancements in AI and hardware, such as robots, are being explored to automate hazardous tasks, with AI’s learning capability allowing machines to work largely unsupervised and improve their performance incrementally from experience. A UK government-funded programme is investigating the use of robots to replace humans in hazardous environments, such as confined spaces and underwater.
AI and Organisational Memory
AI can also defend against the loss of organisational memory, preventing the dilution or removal of important safety controls by reminding organisations of past incidents and the reasons for adopting specific safety measures.
AI's capacity to synthesise vast amounts of information can make businesses more efficient and safe. It offers unprecedented opportunities to improve occupational health and safety, from advanced surveillance to decision-making in hazardous situations. However, ethical considerations, transparency, and regulation are crucial to mitigate risks and ensure the responsible use of AI in the workplace. The advancements in AI are not just about efficiency and safety; they are about creating a harmonious workplace where technology and humans coexist and complement each other.