Using an Artificial intelligence chatbot to critically review the scientific literature on the use of Artificial intelligence in Environmental Impact Assessment

Figures & data

Table 1. The protocol outputs for each of the issues identified in Box 2 (note that the citations identified all appear in Box 2).

AI Issue	Valid citations	Which step you got to in order to find citations; If step 4, what keywords, in what database (Scopus or Google Scholar) found citation
Data Quality and Availability	Cortés et al. (2000); Nadeem and Hameed (2008); Schibuola and Byer (1991); Wang et al. (2006)	Step 4 (in Elicit search string “What is the evidence that in the context of EIA, poor quality data can lead to incomplete assessments and misinformed decisions”.
Transparency and Interpretability	Curmally et al. (2022); de Fine Licht and de Fine Lichtde Fine Licht (2020); Stahl and Leach (2023); Wischmeyer (2020)	Step 4: in Elicit, used “transparency” AND “black boxes” AND “artificial intelligence” to get Wischmeyer (2020) and also de Fine Licht and de Fine Lichtde Fine Licht (2020). Then in Google Scholar, using “environmental impact assessment” AND “trust” AND “artificial intelligence” to find Stahl and Leach (2023) and also Curmally et al. (2022).
Over-reliance on AI	Broussard (2018); Campolo and Crawford (2020); Khan and Nawaz Chaudhry (2023)	Step 4: in Google Scholar and Scopus using “artificial intelligence” AND “over-reliance” OR “trust” and following citations in papers identified.
Loss of Expert Intuition	Khan and Nawaz Chaudhry (2023); Spector and Shanshan (2019)	Step 4: using Elicit and more general keywords “artificial intelligence” AND “expert input” OR “intuition” OR “critical thinking” and following citations in papers identified.
Ethical Considerations	McGovern et al. (2022)	Step 1.
Data Security and Privacy	Di Minin et al. (2021)	Step 4: using Google Scholar with “data security and privacy” AND “artificial intelligence” AND “conservation”.
Economic and Social Impacts	Aghion et al. (2018); Bonab et al. (2021); Gurstein (1985); Hagerty and Rubinov (2019); Uren and Edwards (2023)	Step 4: using Scopus, Google Scholar and Elicit with “economic and social Impacts” AND “artificial intelligence”.
Regulatory and Legal Issues	Giuffrida (2019); Lozo and Onishchenko (2021); Makhkamov (2022); Pagallo et al. (2022)	Makhkamov (2022) obtained through step 1. Lozo and Onishchenko (2021) and Pagallo et al. (2022) found using step 4 and Elicit with “legal issues” AND “artificial intelligence” AND “environmental impact assessment”. Giuffrida found using step 4 and Elicit and “legal implications of AI use in decision making”.
Standardization and Validation	Curmally et al. (2022); Kaur et al. (2022); McGovern et al. (2022); Schwartz et al. (2022); Um et al. (2022); Koyamparambath et al. (2022)	McGovern et al. (2022) found using Step 1. Schwartz et al. (2022); Kaur et al. (2022); Um et al. (2022) found using Step 3 with Google Scholar. Curmally et al. (2022); Koyamparambath et al. (2022) found using Step 4 using Google Scholar with “data standardization” AND “data validation” AND “artificial intelligence” AND “environmental impact assessment”.
EnvironmentalMisrepresentations	Cortés et al. (2000); Galaz et al. (2021); Alemohammad et al. (2023)	Cortés et al. (2000) found using Step 1. Galaz et al. (2021) found using Step 3 and Google Scholar. Alemohammad et al. (2023) already known to authors (addition outside the Protocol)
Temporal and Spatial Challenges	Cortés et al. (2000); Rizzoli and Young (1997)	Step 4: Google Scholar. Adapted the key words “temporal and spatial challenges” AND “artificial intelligence” AND “environmental impact assessment” to “temporal and spatial challenges of artificial intelligence in environmental impact assessment”.
Scalability and Transferability	Francini et al. (2023); Liu and Chih-Wei (2009); Mustak et al. (2023)	Liu and Chih-Wei (2009 found using Step 1). Other references found using Step 3 and Google Scholar.

Figure 1. AI issues in EIA as identified by ChatGPT, without author interpretation. The inner green circles represent level 1 issues, connection shown to all level 2 factors (yellow rounded boxes) with dashed lines.

Figure 2. AI issues and factors in EIA as identified by ChatGPT, after author interpretation. There are three concentric circles: 1) the inner green circles represent level 1 issues, connection shown to all level 2 factors (yellow rounded boxes) with lines; connection to other level 1 issues have replaced circular arguments to the same issue that had been listed in level 2; 2) the middle yellow rounded boxes are level 2 factors that are connected to more than one level 1 issue using solid lines; 3) the outer yellow rounded boxes are level 2 factors that connect to a single level 1 issue using dashed lines.

Cortés U, Sànchez-Marrè M, Ceccaroni L, R-Roda I, Poch M. 2000. Artificial intelligence and environmental decision support systems. Appl Intell. 13(1):77–91. doi: 10.1023/A:1008331413864.

 Web of Science ®Google Scholar

Nadeem O, Hameed R. 2008. Evaluation of environmental impact assessment system in Pakistan. Environ Impact Assess Rev. 28(8):562–71. doi: 10.1016/j.eiar.2008.02.003.

 Web of Science ®Google Scholar

Schibuola S, Byer PH. 1991. Use of knowledge-based systems for the review of environmental impact assessments. Environ Impact Assess Rev. 11(1):11–27. doi: 10.1016/0195-9255(91)90014-B.

 Google Scholar

Wang YM, Yang JB, Xu DL. 2006. Environmental impact assessment using the evidential reasoning approach. Eur J Oper Res. 174(3):1885–913. doi: 10.1016/j.ejor.2004.09.059.

 Web of Science ®Google Scholar

Curmally A, Sandwidi BW, Jagtiani A. 2022. Chapter 9: Artificial intelligence solutions for environmental and social impact assessments. In: Fonseca A, editor. Handbook of Environmental Impact Assessment. Cheltenham: Edward Elgar Publishing Limited; p. 163–177.

 Google Scholar

de Fine Licht K, de Fine Lichtde Fine Licht J. 2020. Artificial intelligence, transparency, and public decision-making. AI Soc. 35(4):917–926. doi: 10.1007/s00146-020-00960-w.

 Web of Science ®Google Scholar

Stahl BC, Leach T. 2023. Assessing the ethical and social concerns of artificial intelligence in neuroinformatics research: an empirical test of the European Union Assessment List for Trustworthy AI (ALTAI)’. AI Ethics. 3(3):745–67. doi: 10.1007/s43681-022-00201-4.

 Google Scholar

Wischmeyer T. 2020. Artificial intelligence and transparency: opening the black box. In: Wischmeyer T, and Rademacher T, editors. Regulating artificial intelligence. Cham: Springer International Publishing; p. 75–101.

 Google Scholar

Broussard M. 2018. Artificial unintelligence: how computers misunderstand the world. Cambridge MA: MIT Press.

 Google Scholar

Campolo A, Crawford K. 2020. Enchanted determinism: power without responsibility in artificial intelligence. Engag Sci Technol Soc. 6:1–19. doi: 10.17351/ests2020.277.

 Web of Science ®Google Scholar

Khan M, Nawaz Chaudhry M. 2023. Artificial intelligence and the future of impact assessment. SSRN Electron J. Available at SSRN 4519498. doi: 10.2139/ssrn.4519498.

 Google Scholar

Spector JM, Shanshan M. 2019. Inquiry and critical thinking skills for the next generation: from artificial intelligence back to human intelligence. Smart Learn Environ. 6(1):1–11. doi: 10.1186/s40561-019-0088-z.

 Google Scholar

McGovern A, Ebert-Uphoff I, John Gagne D, Bostrom A. 2022. Why we need to focus on developing ethical, responsible, and trustworthy artificial intelligence approaches for environmental science. Environ Data Sci. 1:e6. doi: 10.1017/eds.2022.5.

 Google Scholar

Di Minin E, Fink C, Hausmann A, Kremer J, Kulkarni R. 2021. How to address data privacy concerns when using social media data in conservation science. Conserv Biol. 35:437–446. doi: 10.1111/cobi.13708.

 PubMed Web of Science ®Google Scholar

Aghion P, Jones BF, Jones CI. 2018. Artificial intelligence and economic growth in the economics of artificial intelligence: an agenda. Chicago: University of Chicago Press.

 Google Scholar

Bonab AB, Rudko I, Bellini F. 2021. A review and a proposal about socio-economic impacts of artificial intelligence. In Business Revolution in a Digital Era: 14th International Conference on Business Excellence, ICBE 2020; Bucharest, Romania: Springer p. 251–270.

 Google Scholar

Gurstein M. 1985. Social impacts of selected artificial intelligence applications: the Canadian context. Futures. 17(6):652–71. doi: 10.1016/0016-3287(85)90018-7.

 Web of Science ®Google Scholar

Hagerty A, Rubinov I. 2019. Global AI ethics: a review of the social impacts and ethical implications of artificial intelligence. doi: 10.48550/arXiv.1907.07892.

 Google Scholar

Uren V, Edwards JS. 2023. Technology readiness and the organizational journey towards AI adoption: an empirical study. Int J Inf Manage. 68:102588. doi: 10.1016/j.ijinfomgt.2022.102588.

 Web of Science ®Google Scholar

Giuffrida I. 2019. Liability for AI decision-making: some legal and ethical considerations. Fordham L Rev. 88:439.

 Web of Science ®Google Scholar

Lozo O, Onishchenko O. 2021. The potential role of the artificial intelligence in combating climate change and natural resources management: political, legal and ethical challenges. J Nat Resour. 4(3):111–31. doi: 10.33002/nr2581.6853.040310.

 Google Scholar

Makhkamov D. 2022. Modern trends in regulation of environmental and legal relations: digitalization and artificial intelligence. Am J Pol Sci Law Crim. 3(01):41–46. doi: 10.37547/tajpslc/Volume04Issue01-07.

 Google Scholar

Pagallo U, Ciani Sciolla J, Durante M. 2022. The environmental challenges of AI in EU law: lessons learned from the artificial intelligence act (AIA) with its drawbacks. Transforming Gov. 16(3):359–376. doi: 10.1108/TG-07-2021-0121.

 Google Scholar

Kaur D, Uslu S, Rittichier KJ, Durresi A. 2022. Trustworthy artificial intelligence: a review. ACM Comput Surv. 55(2):1–38. doi: 10.1145/3491209.

 Web of Science ®Google Scholar

Schwartz R, Vassilev A, Greene K, Perine L, Burt A, Hall P. 2022. Towards a standard for identifying and managing bias in artificial intelligence. Gaithersburg, MD: National Institute of Standards and Technology, US Department of Commerce; p. 1270. doi: 10.6028/NIST.SP.1270.

 Google Scholar

Um T-W, Kim J, Lim S, Lee GM. 2022. Trust management for artificial intelligence: A standardization perspective. Appl Sci. 12:6022. doi: 10.3390/app12126022.

 Google Scholar

Koyamparambath A, Adibi N, Szablewski C, Adibi SA, Sonnemann G. 2022. Implementing artificial intelligence techniques to predict environmental impacts: Case of construction products. Sustainability. 14(6):3699. doi: 10.3390/su14063699.

 Web of Science ®Google Scholar

Galaz V, Centeno MA, Callahan PW, Causevic A, Patterson T, Brass I, Baum S, Farber D, Fischer J, Garcia D. 2021. Artificial intelligence, systemic risks, and sustainability. Technol Soc. 67:101741. doi: 10.1016/j.techsoc.2021.101741.

 Web of Science ®Google Scholar

Alemohammad S, Casco-Rodriguez J, Luzi L, Imtiaz Humayun A, Babaei H, LeJeune D, Siahkoohi A, Baraniuk RG. 2023. Self-consuming generative models go mad. doi: 10.48550/arXiv.2307.01850.

 Google Scholar

Rizzoli AE, Young WJ. 1997. Delivering environmental decision support systems: software tools and techniques. Environ Model & Soft. 12(2–3):237–49. doi: 10.1016/S1364-8152(97)00016-9.

 Web of Science ®Google Scholar

Francini M, Salvo C, Vitale A. 2023. Combining deep learning and multi-source GIS methods to analyze urban and greening changes. Sensors. 23(8):3805. doi: 10.3390/s23083805.

 PubMed Web of Science ®Google Scholar

Liu KF-R, Chih-Wei Y. 2009. Integrating case-based and fuzzy reasoning to qualitatively predict risk in an environmental impact assessment review. Environ Model Softwa. 24:1241–51. doi: 10.1016/j.envsoft.2009.04.005. 10

 Web of Science ®Google Scholar

Mustak S, Singh D, Kumar Srivastava P. 2023. Advanced remote sensing for Urban and landscape ecology. Singapore: Springer.

 Google Scholar