Use of AI through the automation of routine tasks such as data entry and reporting reduces administrative burden and standardises documentation processes (Baiocco et al., 2022). Reporting and the automation data management allows for the streamlining of compliance processes and enhancing accountability and consistency. This increases efficiency while enabling occupational hygienists to reallocate time towards interpretation of exposure variability, evaluation of control adequacy and strategic risk assessments. Further, it allows for the standardising of templates and real time access to data for organisations from different geographical locations which traditionally consumes a lot of time (Adikwu et al., 2023). As routine administrative functions become increasingly automated; by logging events, generating detailed reports and providing recommendations, we may likely observe greater accuracy and speed and accuracy of responses. This enables a greater commitment toward analytical judgement and helps reinforce hygienists as the guardrails for worker health and safety (AIHA, 2025).

Technology in the field is also developing at an exciting pace. For example, AI integrated smart sensor systems allow for continuous monitoring of environmental parameters (El-Helaly, 2024), increasing the frequency and continuity of exposure data collection beyond traditional snapshot sampling. Increased data continuity enables greater accuracy when identifying trends and deviations which supports proactive intervention and risk management. The incorporation of smart programs can also capture data that might typically get missed by people (El-Helaly, 2024). This, however, doesn’t come without its challenges. The increased volume of real time data requires specific attention paid to its validation, calibration and interpretation to ensure that findings translate into reliable decision making.

Further, AI can use historic exposure data to carry out predictive risk scenario modelling and simulate the effectiveness of control measures at the design stage (Shkembi et al., 2026). This supports a transition from reactive monitoring to anticipatory risk mitigation, embedding occupational hygiene considerations earlier within organisational decision making. By implementing this scenario modelling prior to control implementation, AI has the potential to reduce long term exposure risks and improves the effectiveness of investments in controls (Shkembi et al., 2026).

AI systems, while capable of analysing large quantities of information, are dependent on the quality, representativeness and assumptions embedded in the training data and model (El-Helaly, 2024). Where training data is incomplete, structurally biased and/or unrepresentative of dynamic workplace conditions, AI-generated outputs can produce blind spots, false precision and lead to health and safety hazards (El-Helaly, 2024). In the occupational hygiene context, where small datasets and high variability are common and processes are constantly evolving, overreliance on AI can risk reinforcing flawed assumptions rather than identifying emerging hazards. Moreover, automation overreliance can lead to bias resulting in professionals and organisations accepting AI-generated conclusions without necessary critical review. As decision making becomes communicated through dashboards and predictive models, there is a risk that human oversight reduces particularly in areas where output has been historically precise or appears statistically confident. This may result in false reassurance particularly for low frequency, high consequence events that may fall outside the scope of the modelled data.

The integration of AI-generated monitoring and algorithm management systems may introduce psychosocial and behavioural implications that indirectly influence occupational exposure profiles (Schulte et al., 2024; Schulte & Streit, 2025). Continuous surveillance, productivity analytics and wearable tracking technologies can alter worker autonomy and behaviour and potentially increase stress, fatigue and/or risk-taking behaviours (Schulte & Streit, 2025). This can introduce mental health issues in workers resulting from the experience of isolation, loss of autonomy and disconnection (El-Helaly, 2024; Baiocco et al., 2022). As the profession of occupational hygiene increasingly intersects with smart monitoring systems, professionals must consider not only the technical validity of the generative AI tools but also their impact on the workplace culture and worker wellbeing.

Like any new technology/proof of concept, there will be teething issues and limitations to what it may be able to achieve. For example, possible limitations to AI systems include data bias, model dependency and contextual blind spots. This necessitates an evolution in professional responsibility rather than a reduction or displacement (El-Helaly, 2024). As novel tools increasingly manage data processing and exposure analytics, the occupational hygienist’s role may move along the spectrum of responsibilities. This could mean less reliance on technical specialism shifting focus toward strategy, governance, and assurance. In practice, data management involves ensuring data is representative, investigating variability, validating sensor calibration and model assumptions. Further, this requires exercising professional judgement to determine the reliability of information used to inform compliance decisions. This evolved role extends beyond technical verification. It requires statistical literacy, process understanding and contextual awareness of how work is performed, alongside an understanding of artificial intelligence systems (AIHA, 2025). Where AI systems may generate statistically confident outputs, the hygienist must determine whether those outputs are sufficiently robust to support defensible risk management decisions. In this context, professional competence is increasingly defined not by data collection alone, but by the capacity to validate, challenge and contextualise AI-generated analysis (Park & Seunghon, 2023). For example, HSE’s Estimation and Assessment of Substance Exposure (EASE) was an early computer-based model that used a series of preset criteria to assess occupational exposure to certain substances. While useful as a screening tool, it benefitted from requiring validation by occupational hygienists when applying the results to real workplace situations (El-Helaly, 2024; Tickner et al., 2005).

The integration of AI into workplace risk management does not displace legal accountability from employers or competent professionals. While AI systems may generate exposure analyses, compliance comparisons, or predictive outputs, responsibility for ensuring that risk has been reduced so far as is reasonably practicable remains with duty holders (HSE, 2026). In this context, the occupational hygienist takes a critical role in interrogating the assumptions supporting AI-generated conclusions and ensuring that decisions are legally defensible. Defensibility extends beyond numerical compliance, it requires demonstrating that data are representative, uncertainties are considered, and control measures are proportionate to the level of risk. Where AI outputs are relied upon without sufficient scrutiny, organisations may expose themselves to regulatory challenges, particularly if harm occurs. Consequently, professional judgement and documented oversight become central to maintaining credibility and meeting statutory obligations in an AI-assisted environment (AIHA, 2025).

The integration of AI is also likely to contribute to stratification within the profession of occupational hygiene. As routine administrative and analytical tasks become automated, the economic value of purely technical or template-driven functions may decline. Occupational hygienists whose practice is limited to data collection and standardised reporting may find their roles compressed by automation. However, their expertise may remain complementary to these novel methods of working allowing intervention when systems fail, fine tuning model parameters and providing flexibility in unforeseen situations that AI currently does not accommodate (Baiocco et al., 2022). In contrast, those capable of systems-level analysis, governance, and strategic risk interpretation become increasingly valuable and differentiated.

This stratification may widen the gap between entry-level technical talent and higher-order professional judgement, reinforcing the importance of advanced competence, statistical literacy, and interdisciplinary understanding. Over time, the profession may evolve toward a model in which foundational tasks are supported by automated systems, with junior technicians providing complementary services while senior hygienists can focus on validation, uncertainty management, stakeholder engagement, and organisational risk leadership. Such evolution does not imply obsolescence, but rather a redistribution of professional value toward complex and judgement-intensive domains (Baiocco et al., 2022).

Nevertheless, there are risks to the profession from poorly implemented AI integration such as the potential of progressive deskilling across the profession. As automated systems assume responsibility for data analysis, compliance comparison, and interpretive drafting, hygienists may engage less frequently in the underlying cognitive processes that build expertise (Braverman, 1974; AIHA, 2025). Over time, reduced exposure to critical evaluation, manual modelling, and uncertainty analysis may contribute to the weakened analytical skills, whereby professional judgement may become increasingly dependent on AI outputs rather than independent reasoning (Schulte & Streit, 2025). This erosion of analytical depth can create challenges for training and competence development. If emerging specialists rely heavily on AI-generated conclusions without fully understanding sampling design, variability, and contextual confounders, foundational expertise may weaken. In high-consequence or novel scenarios, where AI models may be least reliable, the absence of deeply developed professional judgement could delay recognition of emerging hazards (NiFhaolain, L. et al., 2023).

Further, overreliance on automated systems introduces reputational and ethical risks (Holweg et al., 2022). Where AI derived conclusions are accepted without rigorous scrutiny, organisations may appear to prioritise technological efficiency over independent oversight. Should harm occur, the profession may face criticism and potential liability, for failing to have the necessary guardrails around AI-usage in occupational hygiene work. The preservation and enablement of independent judgement, therefore, becomes central to maintaining public trust and ongoing professional credibility.

AI has the capacity to significantly enhance occupational hygiene through improved monitoring, predictive risk modelling, and the automation of routine analytical tasks, which frees up time for organisations to proactively address hazards and maintain compliance (Adikwu et al., 2023). In doing so, it shifts the profession away from measurement-heavy technical execution towards governance, validation, and strategic risk leadership. However, the same systems that increase efficiency also have limitations, including data bias, automation overreliance, and psychosocial implications that may alter exposure dynamics in various ways.

The future of occupational hygiene will not be defined by whether AI is adopted, but by how it is governed. As certain functions become automated, the profession is likely to stratify, placing greater value on those capable of interrogating uncertainty, validating outputs, and exercising independent judgement in legally and ethically accountable contexts. AI does not eliminate the need for occupational hygienists; rather, it elevates the importance of critical oversight and professional integrity. The long-term success of AI in occupational hygiene will depend not on its technical sophistication, but on the preservation of competence, accountability, and governance in the protection of worker health.

REFERENCES

Adikwu, F. E., Ozobu, C. O., Odujobi, O., Onyeke, F. O., & Nwulu, E. O. (2023). Advances in EHS Compliance: A Conceptual Model for Standardizing Health, Safety, and Hygiene Programs Across Multinational Corporations. Iconic Research And Engineering Journals, 7(5). https://www.irejournals.com/formatedpaper/1707046.pdf

American Industrial Hygiene Association. (2025). Using AI to enhance OEHS. AIHA. https://aiha-assets.sfo2.digitaloceanspaces.com/AIHA/resources/Guidance-Documents/Guidance-Document-Using-AI-to-Enhance-OEHS.pdf

Baiocco, S., Fernandez-Macías, E., Rani, U., & Pesole,  A. (2022). The role of artificial intelligence in the world of work. International Labour Organization & European Commision. https://www.ilo.org/sites/default/files/wcmsp5/groups/public/%40ed_emp/documents/publication/wcms_849220.pdf

Braverman, H. (1974) Labor and Monopoly Capital: The Degradation of Work in the Twentieth Century. Monthly Review Press, 26.

https://doi.org/10.14452/MR-026-03-1974-07_1

El-Helaly, M. (2024). Artificial intelligence and occupational health and safety: benefits and drawbacks. La Medicina del Lavoro, 115(2), e2024014. https://doi.org/10.23749/mdl.v115i2.15835

Holweg, M., Younger, R., & Wen, Y. (2022). The Reputational Risks of AI. California Management Review Insights. https://cmr.berkeley.edu/2022/01/the-reputational-risks-of-ai/

NiFhaolain, L., Hines, A., & Nallur, V. (2023). Statutory professions in AI governance and their consequences for explainable AI. https://arxiv.org/pdf/2306.08959

Health and Safety Executive. (2026). HSE’s regulatory approach to artificial intelligence (AI). https://www.hse.gov.uk/news/hse-ai.htm

Park, J.H., & Seunghon, H. (2023). Applications and Concerns of Generative AI: ChatGPT in the Field of Occupational Health. Journal of Korean Society of Occupational and Environmental Hygiene, 33(4), 412–418. https://doi.org/10.15269/JKSOEH.2023.33.4.412

Schulte, P. A., & Streit, J. M. K. (2025). Psychosocial risks and ethical implications of technology: considerations for decent work. Annals of Work Exposures and Health, 69(4). https://doi.org/10.1093/annweh/wxaf003

Schulte, P. A., Sauter, S. L., Pandalai, S. P., Tiesman, H. M., Chosewood, L. C., Cunningham, T. R., Wurzelbacher, S. J., Pana‐Cryan, R., Swanson, N. G., Chang, C., Jeannie, Reissman, D. B., Ray, T. K., & Howard, J. (2024). An urgent call to address work‐related psychosocial hazards and improve worker well‐being. American Journal of Industrial Medicine, 67(6). https://doi.org/10.1002/ajim.23583

Shkembi, A., Virji, M. A., He, J., Nambunmee, K., Ruiz-Rudolph, P., & Neitzel, R. L. (2026). Using casual inference and machine learning with exposure determinant modeling to identify important workplace controls. Annals of Work Exposures and Health, 70(1). https://doi.org/10.1093/annweh/wxaf069

Tickner, J., Friar, J., Creely, K.S., Cherrie, J.W., Pryde, D.E., & Kingston, J. (2005). The development of the EASE model. The Annals  of Occupational Hygiene, 49(2), 103-110. https://doi: 10.1093/annhyg/meh085