View in EDS HTML Full Text PDF Full Text

Introducing Generative Artificial Intelligence into the MSW Curriculum: A Proposal for the 2029 Educational Policy and Accreditation Standards

Saved in:
Bibliographic Details
Title: Introducing Generative Artificial Intelligence into the MSW Curriculum: A Proposal for the 2029 Educational Policy and Accreditation Standards
Language: English
Authors: Maria Y. Rodriguez (ORCID 0000-0003-1401-2099), Lauri Goldkind (ORCID 0000-0002-0967-3960), Bryan G. Victor (ORCID 0000-0002-2092-912X), Barbara Hiltz, Brian E. Perron
Source: Journal of Social Work Education. 2024 60(2):174-182.
Availability: Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals
Peer Reviewed: Y
Page Count: 9
Publication Date: 2024
Document Type: Journal Articles
Reports - Evaluative
Descriptors: Artificial Intelligence, Intelligent Tutoring Systems, Educational Policy, Social Work, Technology Uses in Education, Minimum Competencies, Curriculum Design, Educational Change, Opportunities, Professional Education
DOI: 10.1080/10437797.2024.2340931
ISSN: 1043-7797
2163-5811
Abstract: The most recent Council on Social Work Education's Educational Policy and Accreditation Standards (EPAS) demands that social workers develop competence in the ethical and professional deployment of technology. Arguably, artificial intelligence has become a critical element in the technological landscape, most recently with the advent of Generative Artificial Intelligence (GenAI). Beginning in late 2022, there has been an explosion of interest in GenAI, along with a massive and ongoing rollout of GenAI tools for personal and professional use, such as ChatGPT. While GenAI will undoubtedly affect social work practice, scholars and ethicists have raised crucial concerns about GenAI, its potential abuses, and misuses, making it critical that social workers are trained in the proper usage of these technologies. Accordingly, we call here for a 10th competency to be added to the 2029 EPAS: Competency Ten: Social Workers demonstrate the knowledge, skills, and understanding to responsibly and effectively use Generative Artificial Intelligence tools. The current article discusses the recent developments in GenAI and offers social work educators guidance for including GenAI content in social work curricula to meet this proposed standard.
Abstractor: As Provided
Entry Date: 2024
Accession Number: EJ1430717
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwHVVvfv2INoTCJVeR9BTrtRAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDBb0DNq9mE7gCP_bKAIBEICBmwZgm3ITR8MLJBq6BljB3OEu3dlZ4ryv1gjIibLMCdIS58ihVtQdESuYf80uRJ7Z7yqfXTAWXVZrneOeeJHLM8sEv6rsHy6hODUjU9JX7O597exjye9LWjR-7Ov8IdkgVCrE7_xhccDkfZ5cBTkYmOdIsPPfKx0Qqkka3DezgocyUOjumKRL4-zlXyz_kLrNsVN3bH_DbgEvoKqq
Text:
  Availability: 1
  Value: <anid>AN0178359284;swe01apr.24;2024Jul12.05:53;v2.2.500</anid> <title id="AN0178359284-1">Introducing Generative Artificial Intelligence Into the MSW Curriculum: A Proposal for the 2029 Educational Policy and Accreditation Standards </title> <p>The most recent Council on Social Work Education's Educational Policy and Accreditation Standards (EPAS) demands that social workers develop competence in the ethical and professional deployment of technology. Arguably, artificial intelligence has become a critical element in the technological landscape, most recently with the advent of Generative Artificial Intelligence (GenAI). Beginning in late 2022, there has been an explosion of interest in GenAI, along with a massive and ongoing rollout of GenAI tools for personal and professional use, such as ChatGPT. While GenAI will undoubtedly affect social work practice, scholars and ethicists have raised crucial concerns about GenAI, its potential abuses, and misuses, making it critical that social workers are trained in the proper usage of these technologies. Accordingly, we call here for a 10th competency to be added to the 2029 EPAS: Competency Ten: Social Workers demonstrate the knowledge, skills, and understanding to responsibly and effectively use Generative Artificial Intelligence tools. The current article discusses the recent developments in GenAI and offers social work educators guidance for including GenAI content in social work curricula to meet this proposed standard.</p> <p>Recent developments in artificial intelligence (AI), including large language models (LLMs) (e.g., ChatGPT, Claude, and Bard) and image-creating models (e.g., Midjourney, DALL·E, and Stable Diffusion), are producing new opportunities and challenges for social work education and practice. Much of the popular discussion about these generative artificial intelligence (GenAI) systems in academia has focused on challenges, especially regarding the implications of AI-generated text and images for academic integrity (see, e.g., Bearman & Ajjawi, [<reflink idref="bib4" id="ref1">4</reflink>]). At the same time, social workers are now practicing in a world transformed by GenAI, and must be prepared to engage with these tools effectively and ethically. The Council on Social Work Education (CSWE) currently charges social workers to develop competence in the ethical deployment of technology, but there is less clarity on how to do so in such a rapidly changing landscape, to say nothing of the general challenges of implementing broadly stated curricular goals at a scale recognizing the disruptive and evolving nature of GenAI (Danneels, [<reflink idref="bib9" id="ref2">9</reflink>]; Felten et al., [<reflink idref="bib13" id="ref3">13</reflink>]).</p> <p>Building on Singer et al.'s ([<reflink idref="bib32" id="ref4">32</reflink>]) work detailing how GenAI could support Master of Social Work (MSW) instruction, this article recommends the establishment of a new Educational Policy and Accreditation Standard (EPAS) competency indicator: <emph>Social workers demonstrate the knowledge, skills, and understanding to responsibly use generative artificial intelligence tools</emph>. We also detail four skills to be developed in the social work curricula to achieve this new competency. Below is a version of the proposed competency, which we hope will serve as a starting point for discussions on the role of social work educators in preparing students for practice in an AI-informed world (see Figure 1).</p> <p>Graph: Figure 1. Competency.</p> <p>The structure of this article comprises four distinct sections. The first provides a general overview of AI, differentiating between older, narrow technologies and the new class of GenAI tools. The second highlights the value and potential of GenAI in the context of social work education and practice. The third section tempers our optimism with critical considerations that acknowledge potential harms and challenges of GenAI in education and practice. We conclude with a description of the four skills we argue would promote effective engagement with GenAI tools in social work practice.</p> <hd id="AN0178359284-2">Artificial intelligence</hd> <p>Broadly, AI refers to computer systems or models capable of mimicking tasks commonly requiring human intelligence. These include, but are not limited to, interpreting/generating language, recognizing patterns, solving problems, and making recommendations. More traditional AI systems, referred to with terms like general/weak or narrow AI (Andrason, [<reflink idref="bib1" id="ref5">1</reflink>]), are rule-based methods that perform specific actions using predefined algorithms and explicit programming, such as <emph>if–then</emph> rules used in decision-making systems. Narrow AI systems have been deployed in social work practice for over a decade, including predictive analytics to assist in clinical decision making (Cariceo et al., [<reflink idref="bib6" id="ref6">6</reflink>]) and simulation-based training for workers and clients (Asakura et al., [<reflink idref="bib2" id="ref7">2</reflink>]; Goldkind, [<reflink idref="bib16" id="ref8">16</reflink>]).</p> <p>AI has evolved significantly with the introduction of machine learning, where algorithms learn from data rather than following preset rules. This evolution has led to more adaptable and effective AI applications across various fields. In social work, for instance, machine learning models have been applied to analyze and organize large volumes of unstructured text documents (Perron et al., [<reflink idref="bib30" id="ref9">30</reflink>]; Victor et al., [<reflink idref="bib33" id="ref10">33</reflink>]).</p> <p>In contrast, GenAI systems use vast amounts of training data to create images, texts, music, video, and other forms of design with results that are often similar or sometimes indistinguishable from human content creators. Literally speaking, GenAI is proliferate in nature: its purpose is to produce a new artifact or object and it does so by learning the underlying pattern of art, text, or whatever media is in question. That is, although GenAI creates new artifacts, it does so by classifying or predicting the next "logical" thing based on the patterns it has learned from the media it has already consumed.</p> <p>LLMs fall within this broader umbrella of GenAI and are arguably the most relevant to social work among the various types of GenAI models. LLMs are AI systems trained on extensive amounts of text data. They can perform a wide range of text-related tasks, including (but not limited to) classification, summarization, generation, and question–answer based on patterns they learned from their underlying training data. Social work students and practitioners can use plain language to interact with GenAI systems. Yet GenAI systems like ChatGPT are built on computer programming—meaning humans made decisions about the data used to train the model, the features to be emphasized in the data, and the ways output would be generated.</p> <hd id="AN0178359284-3">Opportunities of GenAI in social work practice</hd> <p>Like its influence in other disciplines and professions, GenAI holds some promise for improving social work practice, <emph>primarily through its ability to ease resource constraints for writing-oriented tasks</emph>. Practice at the individual level often includes case management and therapeutic interventions. GenAI has the potential to enhance individual practice as a tool for documentation and reflection (Flemotomos et al., [<reflink idref="bib14" id="ref11">14</reflink>]), as well as a clinical enhancement similar to worksheets one might use in cognitive behavioral therapy. LLMs might be able to help draft treatment plans and case notes, provide materials for both individual and group sessions, assist with translations, and aid in developing other writing-based skills.</p> <p>Many social work organizations face resource constraints and uncertainties (Golensky & Mulder, [<reflink idref="bib18" id="ref12">18</reflink>]). GenAI tools offer mezzolevel practitioners and organizations new ways for automating routine, writing-intensive tasks. For example, LLMs can be used to create customized training materials and resources for ongoing professional development, ensuring that staff are up to date with the latest practices and organizational policies. Social workers at this level might also use GenAI tools to assist in drafting grant proposals, fundraising letters, and other materials needed for resource development, potentially enhancing the quality and efficiency of these critical tasks.</p> <p>Finally, macro social work is broadly focused on policy-level practice and systems change efforts. As other social work scholars have noted (e.g., McBeath, [<reflink idref="bib25" id="ref13">25</reflink>]), AI technologies can be useful to social work practitioners engaged at this level. For instance, policy analysis and subsequent advocacy efforts rely heavily on the review and production of written materials. LLMs could offer time savings by summarizing long policy documents and drafting advocacy materials tailored to different audiences. GenAI models can also be used to help organizations work more effectively with their own documents, helping organizations better use their existing resources, quickly retrieve and synthesize information from their own documents to answer queries, make decisions, or develop strategies. This can be particularly helpful in large organizations where valuable information may be scattered across numerous sources.</p> <p>These practice application examples are not exhaustive; rather, are meant to demonstrate the flexibility and ubiquity of AI, and especially GenAI applications, for social work practice settings. Social work students have an opportunity to be well-positioned for both implementing these tools and shaping policy around their use. The current higher education environment is unclear about whether and how to use GenAI, resulting in much fear and aversion throughout academia. Yet, without exposure, understanding, and reflection, students may leave graduate MSW programs without the necessary knowledge and competence to successfully engage with the GenAI ecosystem.</p> <hd id="AN0178359284-4">Challenges and potential harms of GenAI</hd> <p>Beyond concerns about plagiarism and academic integrity (Wang, Akins, et al., [<reflink idref="bib35" id="ref14">35</reflink>]), GenAI poses real threats to human and environmental well-being in new and accelerated ways. There is currently a heated debate about regulating GenAI in the United States, as well as globally (Chakravorti, [<reflink idref="bib7" id="ref15">7</reflink>]; Matthews, [<reflink idref="bib24" id="ref16">24</reflink>]). The Biden administration has issued an executive order on "safe, secure, and trustworthy AI" (Klare, [<reflink idref="bib21" id="ref17">21</reflink>]), which summarily calls for AI companies to make public their safety testing results with the U.S. government, gives the National Institute for Standards and Technology regulating leverage over AI model safety testing by commercial red teams,[<reflink idref="bib1" id="ref18">1</reflink>] and sets standards for detecting and labeling AI-generated content (charged to the Department of Commerce), among other noteworthy national-level standards. The same administration has also set forth the "Blueprint for an AI Bill of Rights,"[<reflink idref="bib2" id="ref19">2</reflink>] which attempts to lay out the rights citizens have in a marketplace now driven by automated systems and decisions. These two documents help illustrate the well-known harms that can result from AI applied in nonjudicious ways.</p> <p>Additionally, the output of GenAI models are reflective of the data on which it is trained (Patton et al., [<reflink idref="bib29" id="ref20">29</reflink>]). Thus, bias in GenAI systems is often geared toward women, people of color, gender and sexually minortized individuals and communities, and those at the intersections of these identities (Dhingra et al., [<reflink idref="bib12" id="ref21">12</reflink>]; Whittaker et al., [<reflink idref="bib38" id="ref22">38</reflink>]). For example, a recent study of images generated by Stable Diffusion and DALL·E demonstrated gender bias. When prompted for the terms "CEO" or "director," 97% of the returned images were White men (Heikkilä, [<reflink idref="bib19" id="ref23">19</reflink>]). Omiye et al. ([<reflink idref="bib28" id="ref24">28</reflink>]) also found that LLMs often generate outputs aligned with race-based rather than personalized medicine, a practice that could cause great harm and lead to improper care.</p> <p>Social work students and practitioners must be wary of the potential for data harms when engaging with GenAI tools. To use LLMs and other GenAI utilities, all users must first agree to a Terms of Service agreement. Terms of Service have become a ubiquitous part of obtaining any service or content online (Goldkind & Wolf, [<reflink idref="bib17" id="ref25">17</reflink>]). Infamous for their impenetrable language and legalese, these contracts offer their users terms that may be unfair or even abusive (De Rosnay, [<reflink idref="bib11" id="ref26">11</reflink>]). In the case of prompt- or question-driven GenAI tools, any data that the user enters into a system can potentially be accessed by the model developer for subsequent training and model development. For example, if a student wishes to summarize a client's case note using ChatGPT, the data could later be accessed and used by OpenAI, the model's developer.</p> <p>Understanding how LLMs are trained and operated is crucial to mitigate the risks of "hallucinations," a term used in computer science and communications to describe errors, mismatches, and omissions by GenAI models (Beutel et al., [<reflink idref="bib5" id="ref27">5</reflink>]). In GenAI image literature, hallucination occurs when models generate inconsistent or absent objects in descriptions or captions of images (Li et al., [<reflink idref="bib23" id="ref28">23</reflink>]). LLMs are similarly susceptible to such errors, often confidently presenting false information. For instance, LLMs might create plausible-sounding academic literature titles based on their training data, even though these articles do not exist (Wu & Dang, [<reflink idref="bib40" id="ref29">40</reflink>]). A strong foundational knowledge in LLMs' training and functionality can help in identifying and reducing the occurrence of these inaccuracies.</p> <hd id="AN0178359284-5">Four skills needed to develop GenAI competency</hd> <p>Social work education in the United States is governed by the CSWE. CSWE creates and monitors the EPAS, which dictate the competencies and broad skills that all social work students should be able to demonstrate upon graduation from an accredited school of social work.</p> <p>Technology features prominently in the first of the nine EPAS competencies put forward by CSWE in 2022. According to CSWE ([<reflink idref="bib8" id="ref30">8</reflink>]), the ability to <emph>demonstrate ethical and professional behavior</emph> (Competency 1) requires that "social workers understand digital technology and the ethical use of technology in social work practice" (p. 8). The EPAS goes on to note that competency is evidenced by the ability to "use technology ethically and appropriately to facilitate practice outcomes" (p. 9). Based on this guidance, both technology-related knowledge and skill are required for competent practice. AI knowledge and skills are centrally critical to achieving the goals for competence articulated in the 2022 EPAS. One mechanism for accomplishing this goal is the introduction of AI content into the MSW curricula (Goldkind, [<reflink idref="bib16" id="ref31">16</reflink>]).</p> <p>We argue there are four foundational skills necessary for competent and effective use of GenAI and LLMs in social work practice, which comprise in turn the assessment basis for our proposed competency. These skills are:</p> <p></p> <ulist> <item> demonstrating conceptual understanding of the ways in which AI models are trained, and the types of tasks they can perform;</item> <p></p> <item> becoming knowledgeable users of GenAI by understanding how to write prompts effectively;</item> <p></p> <item> assessing output from GenAI models for ethical, socioemotional, and practice implications; and</item> <p></p> <item> cultivating intellectual curiosity by continuously engaging in active learning about the benefits and harms of using GenAI tools at the individual, community, organizational, and national level.</item> </ulist> <hd id="AN0178359284-6">Conceptual understanding of AI models</hd> <p>Using GenAI effectively requires a broad knowledge of what constitutes GenAI and its applications as well as whether, when, and how to use them in practice.</p> <hd id="AN0178359284-7">Definitions and terms</hd> <p>Establishing conceptual grounding includes clarifying crucial definitions. This not only demystifies these technologies but also allows for informed and critical engagement in their applications. Programs should take care to provide consistent definitions of key concepts, such as AI, GenAI, and LLMs, with a focus on what distinguishes generative technologies from other forms of AI.</p> <hd id="AN0178359284-8">Model training</hd> <p>Before deployment of GenAI tools, students should understand how these models were developed and what it means to "train" a model. Training GenAI models, such as LLMs, involves feeding vast amounts of data, including text, images, and sometimes sound, into the model so it can learn patterns, language structures, or relationships among the data elements (e.g., words, images). This training process is facilitated through use of machine learning algorithms that gradually fine-tune the model's settings to improve predictions and generate more accurate and contextually relevant responses or content. For instance, in the context of social work we might look to develop a GenAI model to support clinical practice that was trained on a corpus of text data that includes case studies, legal regulations, ethical guidelines, diagnostic criteria, and textbooks, enabling it to provide informed suggestions that account for the context of individual situations. However, it is crucial for students to understand that the quality and diversity of the training data significantly influence the model's outputs, and they should be aware of potential biases or limitations inherent in the data used for training.</p> <hd id="AN0178359284-9">Data security and confidentiality</hd> <p>In addition, social workers regularly engage with sensitive and often confidential information about individuals and communities. Practitioners must ensure that data submitted into third-party GenAI systems like ChatGPT do not include identifying or protected medical information, or that the use of such data is covered by informed consent agreements.</p> <hd id="AN0178359284-10">Computational thinking</hd> <p>Wing ([<reflink idref="bib39" id="ref32">39</reflink>]) noted that "computational thinking involves solving problems, designing systems, and understanding human behavior, by drawing on the concepts fundamental to computer science" (p. 33). That is, computational thinking involves approaching problems in a way that leverages computational concepts. For social work students, this includes the ability to break complex issues into manageable pieces that can be addressed systematically, creating representations of real-world situations that can be analyzed, and identifying patterns and trends within data, which can lead to evidence-informed decision making. Developing curricular resources to enhance MSW students' computational thinking skills can help to ensure that students gain an understanding of the capabilities and limitations of GenAI. For example, case studies or scenarios can highlight situations where GenAI technologies could be applied in social work contexts, and used to evaluate whether using a GenAI solution aligns with ethical standards or promotes social justice.</p> <hd id="AN0178359284-11">Prompting skills</hd> <p>Prompt engineering is the central skill for interacting with GenAI models and getting desired outputs for a given task (Wang, Shi, et al., [<reflink idref="bib34" id="ref33">34</reflink>]). Prompting GenAI models is a wholly different skill than performing web searches with a range of strategies for securing desired outputs. Techniques in prompt engineering are rapidly evolving, with considerable research showing that slightly reformulated prompts can significantly affect and improve results. For example, chain of thought prompting instructs the LLM to explain its reasoning when answering a prompt, which can lead to significant gains in accuracy (Wei et al., [<reflink idref="bib36" id="ref34">36</reflink>]). White et al. ([<reflink idref="bib37" id="ref35">37</reflink>]) have introduced a catalog detailing various methods for designing prompts. This catalog features techniques like personas, game play, flipped interactions, and context managers. These methods can be customized to address challenges in diverse areas and for the variety of professional tasks that social workers complete across practice levels.</p> <hd id="AN0178359284-12">Assessing model outputs</hd> <p>While GenAI tools can be an integral tool for practitioners, their utility will only be as high as the quality of their output. As noted previously, GenAI outputs can often include hallucinations (i.e., inaccuracies) for which users must assess and correct prior to using any text or code produced by these tools. This process of assessing the quality and accuracy of model outputs is often referred to as output validation (Bandi et al., [<reflink idref="bib3" id="ref36">3</reflink>]). The tasks of prompting and validating are often iterative, as the detection of inaccuracies, or an appraisal that the output was not of sufficient quality for the task at hand, leads to a new round of prompting with necessary adjustments. Additionally, as noted previously, GenAI models are known to reflect the bias that is present in the training data. Social work courses should investigate how outputs generated by these systems could perpetuate this bias and what mitigation is possible.</p> <hd id="AN0178359284-13">Cultivating intellectual curiosity</hd> <p>It is incumbent on MSW programs to safeguard the ethical value of competence by providing mechanisms for MSW students to be lifelong learners of new technologies (Silva et al., [<reflink idref="bib31" id="ref37">31</reflink>]). This is also part of Competency 1 in the current EPAS: "Social workers recognize the importance of lifelong learning and are committed to continually updating their skills to ensure relevant and effective practice."[<reflink idref="bib3" id="ref38">3</reflink>] Lifelong learning is also a value found in the National Association of Social Workers ([<reflink idref="bib27" id="ref39">27</reflink>]) <emph>Code of Ethics</emph>, under the competence principle, "Social workers continually strive to increase their professional knowledge and skills and to apply them in practice. Social workers should aspire to contribute to the knowledge base of the profession."</p> <p>The pace of GenAI technology's evolution is outstripping the current capacity for adaptation in social work curricula. For instance, various applications now enable the installation of small-scale LLMs that operate locally and offline. This development presents new opportunities to address security concerns related to handling sensitive data. Retrieval-augmented generation (RAG), a recent development in GenAI technologies (Gao et al., [<reflink idref="bib15" id="ref40">15</reflink>]; Lewis et al., [<reflink idref="bib22" id="ref41">22</reflink>]), is particularly applicable to social work. RAG improves the responses of LLMs by incorporating data from an external document database, allowing organizations and individuals to interact with their local knowledge. This method reduces inaccuracies and ensures responses are contextually appropriate. Although the most sophisticated RAG methods require programming skills, increasingly available no-code tools like Azure Machine Learning (Microsoft, [<reflink idref="bib26" id="ref42">26</reflink>]) are making this technology accessible to individuals without technical expertise.</p> <p>Accordingly, social work programs need support to know what to teach and how, so they can develop the expertise needed within their schools and departments to support students in achieving this competency prior to graduation. Continuing education opportunities will also be paramount for social work practitioners to maintain and further develop their competence with GenAI tools as these technologies continue to evolve. Collaborations with other departments in the university or private sector practitioners might prove particularly useful in ensuring ongoing training opportunities for alumni.</p> <hd id="AN0178359284-14">Conclusion</hd> <p>Incorporating content on GenAI technologies into the MSW curriculum is not merely a response to technological trends; it is a proactive step toward empowering social workers with the tools and insights required to excel in this new landscape. By cultivating a strong foundation in fundamental knowledge, practical applications, and ethical awareness, social workers will be able to confidently navigate the integration of GenAI into practice, ultimately enhancing their ability to support the individuals and communities they serve.</p> <p>There are two primary strategies for introducing new content into postsecondary curricular offerings: <emph>infusion</emph> of content across existing areas or <emph>block insertions</emph> of specialized courses (De Jong & Naranjo, [<reflink idref="bib10" id="ref43">10</reflink>]). Block insertions are the siloing of new content on populations, practice areas, or innovations into one course, whereby the remaining curricula remains unchanged. Infusing new content is the notion of including a specific key conceptual area, such as GenAI, across relevant courses. The infusion model ensures that all students have multiple touchpoints with new content. An infusion approach is often attractive, as it tends to work within existing curricular structures and hence may be easier and more cost-effective to implement than the insertion of an additional course or courses (De Jong & Naranjo, [<reflink idref="bib10" id="ref44">10</reflink>]; Jin et al., [<reflink idref="bib20" id="ref45">20</reflink>]). We are also mindful that an infusion approach is not without its challenges, as it requires training or hiring faculty with the requisite skills and additional oversight on the part of administrators to ensure effective implementation.</p> <p>While we believe that the integration of this knowledge is critical to future social workers, we are not naive to the multitude of challenges presented by the task of integrating new content into a curriculum. Integrating new content necessitates the allocation of resources—both in terms of time and funding. The dynamic nature of technology requires careful consideration of how faculty will stay abreast of the changing landscape. Balancing the inclusion of new curricular content demands thoughtful planning to ensure cohesion and consistency, while avoiding unnecessary duplication. Furthermore, variations in teaching styles and approaches might inadvertently result in fragmented learning experiences for students.</p> <p>Addressing these complexities will involve collaborative efforts among faculty, administrators, and stakeholders. It will require a strategic approach that considers curriculum design, faculty development, resource allocation, and ongoing program evaluation. Such a comprehensive approach will allow social work programs to harness the opportunities of GenAI to enhance knowledge and skills while upholding the core values of the social work profession.</p> <hd id="AN0178359284-15">Disclosure statement</hd> <p>No potential conflict of interest was reported by the author(s).</p> <ref id="AN0178359284-16"> <title> References </title> <blist> <bibl id="bib1" idref="ref5" type="bt">1</bibl> <bibtext> Andrason, S. P. (2020). Performance of an AGI-aspiring system & narrow-AI approaches: A systematic comparison [ Doctoral dissertation ]. Reykjavík Univeristy.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref7" type="bt">2</bibl> <bibtext> Asakura, K., Occhiuto, K., Todd, S., Leithead, C., & Clapperton, R. (2020). A call to action on artificial intelligence and social work education: Lessons learned from a simulation project using natural language processing. Journal of Teaching in Social Work, 40 (5), 501 – 518. https://doi.org/10.1080/08841233.2020.1813234</bibtext> </blist> <blist> <bibl id="bib3" idref="ref36" type="bt">3</bibl> <bibtext> Bandi, A., Adapa, P. V. S. R., & Kuchi, Y. E. V. P. K. (2023). The power of generative AI: A review of requirements, models, input–output formats, evaluation metrics, and challenges. Future Internet, 15 (8), 260. https://doi.org/10.3390/fi15080260</bibtext> </blist> <blist> <bibl id="bib4" idref="ref1" type="bt">4</bibl> <bibtext> Bearman, M., & Ajjawi, R. (2023). Learning to work with the black box: Pedagogy for a world with artificial intelligence. British Journal of Educational Technology, 54 (5), 1160 – 1173. https://doi.org/10.1111/bjet.13337</bibtext> </blist> <blist> <bibl id="bib5" idref="ref27" type="bt">5</bibl> <bibtext> Beutel, G., Geerits, E., & Kielstein, J. T. (2023). Artificial hallucination: GPT on LSD? Jornal of Critical Care, 27 (1), 148. https://doi.org/10.1186/s13054-023-04425-6</bibtext> </blist> <blist> <bibl id="bib6" idref="ref6" type="bt">6</bibl> <bibtext> Cariceo, O., Nair, M., & Lytton, J. (2018). Data science for social work practice. Methodological Innovations, 11 (3), 1 – 8. https://doi.org/10.1177/2059799118814392</bibtext> </blist> <blist> <bibl id="bib7" idref="ref15" type="bt">7</bibl> <bibtext> Chakravorti, B. (2023, August 4). The AI regulation paradox. Foreign Policy. https://foreignpolicy.com/2023/08/04/ai-regulation-artificial-intelligence-democracy-elections/</bibtext> </blist> <blist> <bibl id="bib8" idref="ref30" type="bt">8</bibl> <bibtext> Council on Social Work Education. (2022). 2022 educational policy and accreditation standards for baccalaureate and master's social work programs. https://<ulink href="http://www.cswe.org/getmedia/bb5d8afe-7680-42dc-a332-a6e6103f4998/2022-EPAS.pdf">www.cswe.org/getmedia/bb5d8afe-7680-42dc-a332-a6e6103f4998/2022-EPAS.pdf</ulink></bibtext> </blist> <blist> <bibl id="bib9" idref="ref2" type="bt">9</bibl> <bibtext> Danneels, E. (2004). Disruptive technology reconsidered: A critique and research agenda. Journal of Product Innovation Management, 21 (4), 246 – 258. https://doi.org/10.1111/j.0737-6782.2004.00076.x</bibtext> </blist> <blist> <bibtext> de Jong, E., & Naranjo, C. (2019). General education teacher educators and English language learner teacher preparation: Infusion as curricular change. The New Educator, 15 (4), 331 – 354. https://doi.org/10.1080/1547688X.2019.1663331</bibtext> </blist> <blist> <bibtext> de Rosnay, M. D. (2016). Alternative policies for alternative internets. Journal of Peer Production. https://halshs.archives-ouvertes.fr/halshs-01362482/document.</bibtext> </blist> <blist> <bibtext> Dhingra, H., Jayashanker, P., Moghe, S., & Strubell, E. (2023). Queer people are people first: Deconstructing sexual identity stereotypes in large language models. arXiv. https://arxiv.org/abs/2307.00101</bibtext> </blist> <blist> <bibtext> Felten, E., Raj, M., & Seamans, R. (2023). How will language modelers like ChatGPT affect occupations and industries? [ Working paper ]. https://arxiv.org/abs/2303.01157</bibtext> </blist> <blist> <bibtext> Flemotomos, N., Martinez, V. R., Chen, Z., Singla, K., Ardulov, V., Peri, R., Caperton, D. D., Gibson, J., Tanana, M. J., Georgiou, P., Epps, J. V., Lord, S. P., Hirsch, T., Imel, Z. E., Atkins, D. C., & Narayanan, S. (2021). Am I a good therapist? Automated evaluation of psychotherapy skills using speech and language technologies. CoRR, Abs, 2102 (10.3758). https://doi.org/10.3758/s13428-021-01623-4</bibtext> </blist> <blist> <bibtext> Gao, Y., Xiong, Y., Gao, X., Jia, K., Pan, J., Bi, Y., Dai, Y., Sun, J., Wang, M., & Wang, H. (2023). Retrieval-augmented generation for large language models: A survey. arXiv preprint arXiv:2312.10997.</bibtext> </blist> <blist> <bibtext> Goldkind, L. (2021). Social work and artificial intelligence: Into the matrix. Social Work, 66 (4), 372 – 374. https://doi.org/10.1093/sw/swab028</bibtext> </blist> <blist> <bibtext> Goldkind, L., & Wolf, L. (2020). Selling your soul on the information superhighway: Consenting to services in direct-to-consumer tele-mental health. Families in Society, 101 (1), 6 – 20. https://doi.org/10.1177/1044389419872125</bibtext> </blist> <blist> <bibtext> Golensky, M., & Mulder, C. A. (2006). Coping in a constrained economy: Survival strategies of nonprofit human service organizations. Administration in Social Work, 30 (3), 5 – 24. https://doi.org/10.1300/J147v30n03_02</bibtext> </blist> <blist> <bibtext> Heikkilä, M. (2023, March 23). These new tools let you see for yourself how biased AI image models are. MIT Technology Review. https://<ulink href="http://www.technologyreview.com/2023/03/22/1070167/these-news-tool-let-you-see-for-yourself-how-biased-ai-image-models-are/">www.technologyreview.com/2023/03/22/1070167/these-news-tool-let-you-see-for-yourself-how-biased-ai-image-models-are/</ulink></bibtext> </blist> <blist> <bibtext> Jin, B., Swinney, J., Cao, H., Muske, G., Nam, J., & Kang, J. (2011). Doing business with China: Curriculum internationalisation through an infusion method. Innovations in Education and Teaching International, 48 (2), 171 – 181. https://doi.org/10.1080/14703297.2011.564012</bibtext> </blist> <blist> <bibtext> Klare, M. T. (2023). Biden issues executive order on AI safety. Arms Control Today, 53 (10), 36 – 37.</bibtext> </blist> <blist> <bibtext> Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., Küttler, H., Lewis, M., Yih, W., Rocktäschel, T., Riedel, S., & Kiela, D. (2020). Retrieval-augmented generation for knowledge-intensive nlp tasks. Advances in Neural Information Processing Systems, 33, 9459 – 9474.</bibtext> </blist> <blist> <bibtext> Li, Y., Du, Y., Zhou, K., Wang, J., Zhao, W. X., & Wen, J. R. (2023). Evaluating object hallucination in large vision-language models. arXiv preprint arXiv:2305.10355.</bibtext> </blist> <blist> <bibtext> Matthews, D. (2023, August 1). The AI rules that US policymakers are considering, explained. Vox. https://<ulink href="http://www.vox.com/future-perfect/23775650/ai-regulation-openai-gpt-anthropic-midjourney-stable">www.vox.com/future-perfect/23775650/ai-regulation-openai-gpt-anthropic-midjourney-stable</ulink></bibtext> </blist> <blist> <bibtext> McBeath, B. (2016). Re-envisioning macro social work practice. Families in Society, 97 (1), 5 – 14. https://doi.org/10.1606/1044-3894.2016.97.9</bibtext> </blist> <blist> <bibtext> Microsoft. (2023, September 30). Use Azure machine learning pipelines with no code to construct RAG pipelines. Microsoft Learn. https://learn.microsoft.com/en-us/azure/machine-learning/how-to-use-pipelines-prompt-flow?view=azureml-api-2</bibtext> </blist> <blist> <bibtext> National Association of Social Workers. (2021). Code of ethics. https://<ulink href="http://www.socialworkers.org/About/Ethics/Code-of-Ethics/Code-of-Ethics-English">www.socialworkers.org/About/Ethics/Code-of-Ethics/Code-of-Ethics-English</ulink></bibtext> </blist> <blist> <bibtext> Omiye, J. A., Lester, J. C., Spichak, S., Rotemberg, V., & Daneshjou, R. (2023). Large language models propagate race-based medicine. NPJ Digital Medicine, 6, 195. https://doi.org/10.1038/s41746-023-00939-z</bibtext> </blist> <blist> <bibtext> Patton, D. U., Landau, A. Y., & Mathiyazhagan, S. (2023). ChatGPT for social work science: Ethical challenges and opportunities. Journal of the Society for Social Work and Research, 14 (3), 553 – 562. https://doi.org/10.1086/726042</bibtext> </blist> <blist> <bibtext> Perron, B. E., Victor, B. G., Bushman, G., Moore, A., Ryan, J. P., Lu, A. J., & Piellusch, E. K. (2019). Detecting substance-related problems in narrative investigation summaries of child abuse and neglect using text mining and machine learning. Child Abuse and Neglect, 98, 104180. https://doi.org/10.1016/j.chiabu.2019.104180</bibtext> </blist> <blist> <bibtext> Silva, D. E., Chen, C., & Zhu, Y. (2022). Facets of algorithmic literacy: Information, experience, and individual factors predict attitudes toward algorithmic systems. New Media & Society, 14614448221098042. https://doi.org/10.1177/14614448221098042</bibtext> </blist> <blist> <bibtext> Singer, J. B., Báez, J. C., & Rios, J. A. (2023). AI creates the message: Integrating AI language learning models into social work education and practice. Journal of Social Work Education, 59 (2), 294 – 302. https://doi.org/10.1080/10437797.2023.2189878</bibtext> </blist> <blist> <bibtext> Victor, B. G., Perron, B. E., Sokol, R. L., Fedina, L., & Ryan, J. P. (2021). Automated identification of domestic violence in written child welfare records: Leveraging text mining and machine learning to enhance social work research and evaluation. Journal of the Society for Social Work and Research, 12 (4), 631 – 655. https://doi.org/10.1086/712734</bibtext> </blist> <blist> <bibtext> Wang, J., Shi, E., Yu, S., Wu, Z., Ma, C., Dai, H., Yang, Q., Kang, Y., Wu, J., Hu, H., Yue, C., Zhang, H., Liu, Y., Pan, Y., Liu, Z., Sun, L., Li, X., Ge, B., Jiang, X., & Zhang, S. (2023). Prompt engineering for healthcare: Methodologies and applications. arXiv preprint arXiv:2304.14670.</bibtext> </blist> <blist> <bibtext> Wang, K., Akins, S., Mohammed, A., & Lawrence, R. (2023). Student mastery or AI deception? Analyzing ChatGPT's assessment proficiency and evaluating detection strategies. arXiv preprint arXiv:2311.16292.</bibtext> </blist> <blist> <bibtext> Wei, J., Wang, X., Schuurmans, D., Bosma, M., Chi, E. H., Xia, F., Le, Q., & Zhou, D. (2022). Chain of thought prompting elicits reasoning in large language models. ArXiv, abs/2201.11903.</bibtext> </blist> <blist> <bibtext> White, J., Fu, Q., Hays, S., Sandborn, M., Olea, C., Gilbert, H., Elnashar, A., Spencer-Smith, J., & Schmidt, D. C. (2023). A prompt pattern catalog to enhance prompt engineering with ChatGPT. ArXiv, abs/2302.11382.</bibtext> </blist> <blist> <bibtext> Whittaker, M., Alper, M., Bennett, C. L., Hendren, S., Kaziunas, L., Mills, M., Morris, M. R., Rankin, J., Emily Rogers, N., Salas, M., & West, S. M. (2019). Disability, bias, and AI. AI Now Institute, 8.</bibtext> </blist> <blist> <bibtext> Wing, J. M. (2006). Computational thinking. Communications of the ACM, 49 (3), 33 – 35. https://doi.org/10.1145/1118178.1118215</bibtext> </blist> <blist> <bibtext> Wu, R. T., & Dang, R. R. (2023). ChatGPT in head and neck scientific writing: A precautionary anecdote. American Journal of Otolaryngology, 44 (6), 103980. https://doi.org/10.1016/j.amjoto.2023.103980</bibtext> </blist> </ref> <ref id="AN0178359284-17"> <title> Footnotes </title> <blist> <bibtext> Red teams work to "break" systems in an attempt to anticipate the actions of adversarial users who would look to exploit a system weakness, as in the case of hackers.</bibtext> </blist> <blist> <bibtext> https://<ulink href="http://www.whitehouse.gov/ostp/ai-bill-of-rights/">www.whitehouse.gov/ostp/ai-bill-of-rights/</ulink>.</bibtext> </blist> <blist> <bibtext> https://<ulink href="http://www.cswe.org/accreditation/standards/2022/">www.cswe.org/accreditation/standards/2022/</ulink>.</bibtext> </blist> </ref> <aug> <p>By Maria Y. Rodriguez; Lauri Goldkind; Bryan G. Victor; Barbara Hiltz and Brian E. Perron</p> <p>Reported by Author; Author; Author; Author; Author</p> <p></p> <p>Maria Y. Rodriguez is Assistant Professor in the School of Social Work and Adjunct Assistant Professor in the Department of Computer Science and Engineering at the University at Buffalo.</p> <p>Lauri Goldkind is Assistant Professor in the Graduate School of Social Service at Fordham University.</p> <p>Bryan G. Victor is Assistant Professor in the School of Social Work at Wayne State University.</p> <p>Barbara Hiltz is Clinical Associate Professor in the School of Social Work at University of Michigan–Ann Arbor.</p> <p>Brian E. Perron is Professor in the School of Social Work at University of Michigan–Ann Arbor.</p> </aug> <nolink nlid="nl1" bibid="bib13" firstref="ref3"></nolink> <nolink nlid="nl2" bibid="bib32" firstref="ref4"></nolink> <nolink nlid="nl3" bibid="bib16" firstref="ref8"></nolink> <nolink nlid="nl4" bibid="bib30" firstref="ref9"></nolink> <nolink nlid="nl5" bibid="bib33" firstref="ref10"></nolink> <nolink nlid="nl6" bibid="bib14" firstref="ref11"></nolink> <nolink nlid="nl7" bibid="bib18" firstref="ref12"></nolink> <nolink nlid="nl8" bibid="bib25" firstref="ref13"></nolink> <nolink nlid="nl9" bibid="bib35" firstref="ref14"></nolink> <nolink nlid="nl10" bibid="bib24" firstref="ref16"></nolink> <nolink nlid="nl11" bibid="bib21" firstref="ref17"></nolink> <nolink nlid="nl12" bibid="bib29" firstref="ref20"></nolink> <nolink nlid="nl13" bibid="bib12" firstref="ref21"></nolink> <nolink nlid="nl14" bibid="bib38" firstref="ref22"></nolink> <nolink nlid="nl15" bibid="bib19" firstref="ref23"></nolink> <nolink nlid="nl16" bibid="bib28" firstref="ref24"></nolink> <nolink nlid="nl17" bibid="bib17" firstref="ref25"></nolink> <nolink nlid="nl18" bibid="bib11" firstref="ref26"></nolink> <nolink nlid="nl19" bibid="bib23" firstref="ref28"></nolink> <nolink nlid="nl20" bibid="bib40" firstref="ref29"></nolink> <nolink nlid="nl21" bibid="bib39" firstref="ref32"></nolink> <nolink nlid="nl22" bibid="bib34" firstref="ref33"></nolink> <nolink nlid="nl23" bibid="bib36" firstref="ref34"></nolink> <nolink nlid="nl24" bibid="bib37" firstref="ref35"></nolink> <nolink nlid="nl25" bibid="bib31" firstref="ref37"></nolink> <nolink nlid="nl26" bibid="bib27" firstref="ref39"></nolink> <nolink nlid="nl27" bibid="bib15" firstref="ref40"></nolink> <nolink nlid="nl28" bibid="bib22" firstref="ref41"></nolink> <nolink nlid="nl29" bibid="bib26" firstref="ref42"></nolink> <nolink nlid="nl30" bibid="bib10" firstref="ref43"></nolink> <nolink nlid="nl31" bibid="bib20" firstref="ref45"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1430717
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Introducing Generative Artificial Intelligence into the MSW Curriculum: A Proposal for the 2029 Educational Policy and Accreditation Standards
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Maria+Y%2E+Rodriguez%22">Maria Y. Rodriguez</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0003-1401-2099">0000-0003-1401-2099</externalLink>)<br /><searchLink fieldCode="AR" term="%22Lauri+Goldkind%22">Lauri Goldkind</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-0967-3960">0000-0002-0967-3960</externalLink>)<br /><searchLink fieldCode="AR" term="%22Bryan+G%2E+Victor%22">Bryan G. Victor</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-2092-912X">0000-0002-2092-912X</externalLink>)<br /><searchLink fieldCode="AR" term="%22Barbara+Hiltz%22">Barbara Hiltz</searchLink><br /><searchLink fieldCode="AR" term="%22Brian+E%2E+Perron%22">Brian E. Perron</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Journal+of+Social+Work+Education%22"><i>Journal of Social Work Education</i></searchLink>. 2024 60(2):174-182.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 9
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2024
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Evaluative
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Artificial+Intelligence%22">Artificial Intelligence</searchLink><br /><searchLink fieldCode="DE" term="%22Intelligent+Tutoring+Systems%22">Intelligent Tutoring Systems</searchLink><br /><searchLink fieldCode="DE" term="%22Educational+Policy%22">Educational Policy</searchLink><br /><searchLink fieldCode="DE" term="%22Social+Work%22">Social Work</searchLink><br /><searchLink fieldCode="DE" term="%22Technology+Uses+in+Education%22">Technology Uses in Education</searchLink><br /><searchLink fieldCode="DE" term="%22Minimum+Competencies%22">Minimum Competencies</searchLink><br /><searchLink fieldCode="DE" term="%22Curriculum+Design%22">Curriculum Design</searchLink><br /><searchLink fieldCode="DE" term="%22Educational+Change%22">Educational Change</searchLink><br /><searchLink fieldCode="DE" term="%22Opportunities%22">Opportunities</searchLink><br /><searchLink fieldCode="DE" term="%22Professional+Education%22">Professional Education</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1080/10437797.2024.2340931
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 1043-7797<br />2163-5811
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: The most recent Council on Social Work Education's Educational Policy and Accreditation Standards (EPAS) demands that social workers develop competence in the ethical and professional deployment of technology. Arguably, artificial intelligence has become a critical element in the technological landscape, most recently with the advent of Generative Artificial Intelligence (GenAI). Beginning in late 2022, there has been an explosion of interest in GenAI, along with a massive and ongoing rollout of GenAI tools for personal and professional use, such as ChatGPT. While GenAI will undoubtedly affect social work practice, scholars and ethicists have raised crucial concerns about GenAI, its potential abuses, and misuses, making it critical that social workers are trained in the proper usage of these technologies. Accordingly, we call here for a 10th competency to be added to the 2029 EPAS: Competency Ten: Social Workers demonstrate the knowledge, skills, and understanding to responsibly and effectively use Generative Artificial Intelligence tools. The current article discusses the recent developments in GenAI and offers social work educators guidance for including GenAI content in social work curricula to meet this proposed standard.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2024
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1430717
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1430717
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1080/10437797.2024.2340931
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 9
        StartPage: 174
    Subjects:
      – SubjectFull: Artificial Intelligence
        Type: general
      – SubjectFull: Intelligent Tutoring Systems
        Type: general
      – SubjectFull: Educational Policy
        Type: general
      – SubjectFull: Social Work
        Type: general
      – SubjectFull: Technology Uses in Education
        Type: general
      – SubjectFull: Minimum Competencies
        Type: general
      – SubjectFull: Curriculum Design
        Type: general
      – SubjectFull: Educational Change
        Type: general
      – SubjectFull: Opportunities
        Type: general
      – SubjectFull: Professional Education
        Type: general
    Titles:
      – TitleFull: Introducing Generative Artificial Intelligence into the MSW Curriculum: A Proposal for the 2029 Educational Policy and Accreditation Standards
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Maria Y. Rodriguez
      – PersonEntity:
          Name:
            NameFull: Lauri Goldkind
      – PersonEntity:
          Name:
            NameFull: Bryan G. Victor
      – PersonEntity:
          Name:
            NameFull: Barbara Hiltz
      – PersonEntity:
          Name:
            NameFull: Brian E. Perron
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 01
              Type: published
              Y: 2024
          Identifiers:
            – Type: issn-print
              Value: 1043-7797
            – Type: issn-electronic
              Value: 2163-5811
          Numbering:
            – Type: volume
              Value: 60
            – Type: issue
              Value: 2
          Titles:
            – TitleFull: Journal of Social Work Education
              Type: main
ResultId 1