What the paper studied
This paper treats AI in hospitality as a socio-technical system rather than a neutral upgrade. Using a qualitative integrative review, the author draws on hospitality management, information systems, organisational studies and AI ethics literature (sourced from Scopus, Web of Science, ScienceDirect, Emerald Insight and Google Scholar, prioritising 2015–2025 publications) and applies thematic synthesis to assess AI's economic, social, ethical, environmental and regulatory implications.
Key findings
- Economic gains are real but unevenly shared. Studies cited in the review report that automated inventory tracking can cut supply costs by up to 30% and that personalised upselling tools can raise room-level revenue by 40–50%. Early adopters build compounding "exponential learning" advantages, widening the gap with smaller operators that lack data, capital or digital skills.
- Work is reallocated more than eliminated. AI absorbs routine tasks while staff shift toward judgement, empathy and problem-solving, but only when automation comes with training and participatory implementation. Without them, it amplifies anxiety about job security, workload, monitoring and fair performance evaluation.
- Personalization creates a trust tension. Guests show a "privacy paradox": they share data when the benefit is clear but object when practices feel opaque or intrusive. Models trained on skewed data can produce unequal outcomes in pricing, recruitment and recommendations, and "black box" systems undermine confidence.
- AI is both a sustainability tool and a footprint. AI energy management can reduce consumption by up to 20% in some cases, and AI monitoring helps kitchens cut food waste. These on-site gains are offset by data centres that consumed an estimated 415 TWh of electricity in 2024 and can use up to two litres of cooling water per kilowatt-hour, impacts that are rarely reported.
- Regulation is fragmented. Privacy laws such as GDPR and CCPA set important safeguards, but guidance on algorithmic fairness, accountability and environmental responsibility is underdeveloped, and few jurisdictions require disclosure of the energy or water use behind cloud services.
- Readiness is a bottleneck. Limited AI knowledge, legacy systems that do not talk to each other (PMS, POS, booking platforms), financial constraints and uneven digital literacy among managers slow adoption.
Why it matters for hospitality
Much of the AI conversation in hospitality is led by efficiency and revenue claims. This paper is a counterweight: it shows how AI can deepen inequality between large groups and independents, shift environmental impact to outsourced infrastructure that sustainability reports leave out, and turn staff into overseers of algorithms rather than hosts. Its central argument is that the industry should ask not only what AI can do, but what it ought to do, and for whom.
Practical takeaways
- Start with phased pilots (for example in demand forecasting, maintenance or personalization) to test systems at a manageable scale and build internal support before full deployment.
- Put governance in place before scaling: data governance, regular bias audits, explainability standards and clear accountability for AI-driven decisions.
- Ask technology providers to report the energy and water use behind their services, and favour energy-efficient options, so digital strategy stays aligned with sustainability commitments.
- Involve employees in implementation and pair automation with reskilling and career pathways to limit resistance and protect autonomy.
- Be clear with guests about what data is collected and why, especially for biometric and behavioural data.
Limitations
This is a single-author conceptual review rather than a systematic review or empirical study. It does not report a formal search protocol or quantitative synthesis, and the cost, revenue and energy figures it presents come from the studies it cites rather than new data.