WorkerBot and business supply chains due diligence: trustworthy digital infrastructure based on LLM
This paper considers the risks associated with a WorkerBot prototype designed to engage with workers, speaking in their native tongue on employment and immigration issues without the fear of repercussions. The WorkerBot exploits all the latest advanced features of LLMs, and hence inherits all the risks posed by LLMs, which can be categorised into data related risks, model complexity-related risks, and those related to social behaviours and human values according to a recent Turing’s report (2024). After analysing the risks associated with the AI WorkerBot, we make recommendations towards a safe adoption. Modern slavery is a century old problem that is hard to eradicate. The potential societal impact of this initiative is profound if the innovation is successful.
Authors
- Eghbal Rahimikia (ORCID: https://orcid.org/0000-0001-6583-2332)
- Ser‐Huang Poon (ORCID: https://orcid.org/0000-0002-7297-9401)
- Philip Jobi Vallavanthra
Institutions
- University of Manchester (GB)
Publication Details
- Journal
- IET conference proceedings.
- Published
- 2025-12-01
- DOI
- https://doi.org/10.1049/icp.2025.1269
- Primary Topic
- AI in Service Interactions
- Type
- article
- Field-Weighted Citation Impact
- 0.00