CoChat
Sign Up
Cited by 335Digital literacy in educationEthics and Social Impacts of AIArtificial Intelligence in Education

Generative AI Literacy: Twelve Defining Competencies

Open access
Annapureddy, Ravinithesh, Fornaroli, Alessandro, Gatica-Perez, Daniel·2025·Digital Government: Research and Practice
DOI 10.1145/3685680
Open this paper in CoChat — free accountOpen PDF

Abstract

This article introduces a competency-based model for generative artificial intelligence (AI) literacy covering essential skills and knowledge areas necessary to interact with generative AI. The competencies range from foundational AI literacy to prompt engineering and programming skills, including ethical and legal considerations. These 12 competencies offer a framework for individuals, policymakers, government officials, and educators looking to navigate and take advantage of the potential of generative AI responsibly. Embedding these competencies into educational programs and professional training initiatives can equip individuals to become responsible and informed users and creators of generative AI. The competencies follow a logical progression and serve as a roadmap for individuals seeking to become familiar with generative AI and for researchers and policymakers to develop assessments, educational programs, guidelines, and regulations.

AI snapshot

Written from the paper’s full text.

TL;DR

The paper proposes a competency-based model for generative AI literacy consisting of twelve competencies, from basic AI literacy to prompt engineering, programming, ethical, and legal knowledge. It argues these competencies can guide education, policy, and professional training for responsible generative AI use.

Key findings

  • The model identifies twelve competencies: basic AI literacy; knowledge of generative AI models; knowledge of capacity and limitations; skill to use generative AI tools; ability to detect AI-generated content; ability to assess output; prompt engineering; ability to program and fine-tune; knowledge of contexts; knowledge of ethical implications; knowledge of legal aspects; and ability to continuously learn.
  • A preliminary database search in December 2023 using “generative AI literacy” OR “generative artificial intelligence literacy” across five databases returned 6 records after duplicates, indicating no well-established concept; existing papers often conflate generative AI literacy with general AI literacy.
  • The paper defines generative AI literacy as a long-term, continuous process of identifying, acquiring, practicing, and mastering core competencies to effectively and responsibly understand, assess, and work with generative AI technologies.
  • It proposes the competencies as a foundation for assessments, curricula, professional development, public-sector guidelines, and further research, while noting that positive and negative implications of each competency are mostly speculative.

Methods

  • Preliminary database search in ACM Digital Library, IEEExplore, Clarivate Web of Science, Scopus, and Dimensions with the query “generative AI literacy” OR “generative artificial intelligence literacy”; after removing duplicates, 6 records were included and summarized.
  • Because of the topic’s novelty, the authors conducted a non-systematic, comprehensive literature review in English covering AI literacy, systematic reviews on generative AI applications, prompt engineering, and competency models.
  • The authors identified twelve competencies from the literature, described each in detail with illustrative examples, and used a speculative approach for discussing potential implications due to scarce prior research.

Limitations

  • The paper states that the novelty of the topic made a long-standing review approach infeasible, that the potential implications are mostly speculative and based on informed opinions and current trends, and that the competencies are designed based on the current state of the field and may need revision as technology advances.
  • The text makes evident that the framework is conceptual and is not empirically validated or tested with a sample; no assessment instrument, proficiency metrics, or outcome data are reported.
  • The preliminary database search was limited to one exact query and five databases, yielding only 6 records, which the authors use to justify the non-systematic review approach.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1145/3685680

Cites

114 of 114 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 58 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    A preliminary database search across five scientific databases using the query 'generative AI literacy' OR 'generative artificial intelligence literacy' resulted in 6 records after removing duplicates.

    “Given the low amount of results (6 records in total, as described in Section 3.1), all the articles are described in Table 1.”
    • Also stated in: § 3.1 Preliminary Database Search · plain
      “After removing duplicates, this search resulted in a total of 6 records, which have been included and summarized in Table 1.”
    Quote in the paperSubstantial claim
  • Premisecitation

    Existing AI literacy frameworks do not differentiate between generative and predictive models.

    “Despite this, the concept of AI literacy has not evolved, and existing frameworks do not differentiate between generative and predictive models [61,62].”
    • Also stated in: § 3.4 Need for a Generative AI Literacy competency model · plain
      “the existing literacy frameworks do not differentiate between them.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [61]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [62]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Premisecitation

    In ChatGPT, adjustable hyperparameters include temperature, maximum number of tokens, top_k, and top_p.

    “In the case of ChatGPT, for example, these include the temperature, which determines the “creativity” of the model, the maximum number of tokens to be used, and the top_k and top_p to limit the number of next word choices of the model and the probability of the next word choices, respectively [60,83].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [60]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [83]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Premisecitation

    The output of ChatGPT and other LLMs can often contain a mixture of facts and completely false and fabricated statements.

    “The output of ChatGPT and other LLMs can often contain a mixture of facts and completely false and fabricated statements, and awareness of this is fundamental [1,5].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [1]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [5]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Premisecitation

    With the advance of generative AI, it has become increasingly difficult to discern which contents have been entirely human-generated, AI-generated, or AI-enhanced.

    “With the advance of generative AI, it has become increasingly difficult to discern which contents have been entirely human-generated, AI-generated, or AI-enhanced [29].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [29]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Classified as probably not a substantial claim. 25
  • May combine two claims. 14

  • Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

    Interpretationreasoning

    The twelve competencies follow a logical progression and serve as a roadmap for individuals seeking to get familiar with generative AI and for researchers and policymakers to develop assessments, educational programs, guidelines, and regulations.

    “The competencies follow a logical progression and serve as a roadmap for individuals seeking to get familiar with generative AI and for researchers and policymakers to develop assessments, educational programs, guidelines, and regulations.”
    • Also stated in: § 4 A COMPETENCY-BASED MODEL FOR GENERATIVE AI LITERACY · plain
      “the ordering of these competencies is purposeful and follows a logical progression to represent a structured learning path reflecting a continuum of learning”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The rapid rise of generative AI tools and their widespread adoption in various disciplines have created a unique set of skills specific to these novel tools, which go beyond the traditional competencies associated with data literacy, digital literacy, or AI literacy.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Generative AI had a disruptive impact on different domains, and not all authorities and organizations have had time to adapt by developing guidelines and rules, as in the scientific and academic world.

    “Generative AI had a disruptive impact on different domains, and not all authorities and organizations have had the time to adapt to it by developing guidelines and rules, as in the case of the scientific and academic world [89,90].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [89]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [90]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Premisecitation

    Some AI detection tools tend to overestimate the probability that a text, image, or video is AI-generated, while others underestimate it, and there is often a great divergence in results across different AI detection tools.

    “some of these detectors tend to overestimate the probability that a text, image, or video is AI-generated, while others underestimate it, and there is often a great divergence in results across different AI detection tools [97].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [97]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    In recent years, there has been a surge in software and AI models to detect AI-generated content, including text, images, and videos.

    “In recent years, there has been a surge in software and AI models to detect AI-generated content, both text (such as ZeroGPTfoot_9 , GPTZerofoot_10 , PlagiarismCheckfoot_11 , GPT Radarfoot_12 , Writer.comfoot_13 , among others) [18,66,97], images (AI or Not 14 or Is it AI? 15 ) and videos [45,75].”
    Quote in the paperNot a substantial claim0 of 5 sources checked
    • cites [18]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [66]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [97]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [45]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [75]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Increasingly more companies require employees who can use and master generative AI.

    “this has become a noteworthy skill to have in the labor market, as increasingly more companies require employees who can use and master generative AI [14,31].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [14]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [31]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Classified as probably not a substantial claim.

    Interpretationnone given

    The twelve competencies identified in this paper can be intended as a complement to the existing framework on AI literacy, providing additional guidance on the specific toolbox that a user of generative AI tools needs.

    “The twelve competencies identified in this paper can be intended as a complement to the existing framework on AI literacy, providing additional guidance on the specific toolbox that a user of generative AI tools needs.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The rapid rise of generative AI tools and their widespread adoption in various disciplines have created a unique set of skills specific to these novel tools, which go beyond the traditional competencies associated with data literacy, digital literacy, or AI literacy.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The concept of generative AI literacy is associated with prompt engineering, which is the ability to design effective prompts for generative AI models to produce desired outputs.

    “This concept of generative AI literacy is associated with that of prompt engineering, which is the ability to design effective prompts for generative AI models to produce desired outputs [79].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [79]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Premisecitation

    Human detection of AI-generated photos and texts becomes increasingly difficult, making AI detectors more significant.

    “However, human detection of AI-generated photos and texts becomes increasingly difficult [68], making AI detectors always more significant.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [68]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Multimodal prompting, including images and other formats as input, has emerged and increased in popularity.

    “Multimodal prompting, including images and other formats as input, has also emerged and increased in popularity [76]”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [76]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Different sets of guidelines have been written on prompt engineering for ChatGPT and LLMs, highlighting its importance.

    “Different sets of guidelines have been written on prompt engineering for ChatGPT and LLMs, highlighting its importance [2,34,87,100].”
    Quote in the paperNot a substantial claim0 of 4 sources checked
    • cites [2]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [34]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [87]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [100]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Interpretationown data

    In the current academic literature, there is not a well-established concept of generative AI literacy; the few existing papers tend to conflate it with the more general concept of AI literacy.

    “As observed from Table 1, there is not a well-established concept of generative AI literacy in the current academic literature.”
    • Also stated in: § 3.1 Preliminary Database Search · plain
      “The few existing papers tend to conflate it with the more general and better-established concept of AI literacy.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    There is a gap in academic literature on generative AI literacy; no set of competencies specific to the use of generative AI had been previously proposed, and no comprehensive framework systematically addresses competency acquisition in navigating and utilizing generative AI tools.

    “This paper aims to fill a gap at the intersection of the topics of AI literacy and generative AI, as no set of competencies specific to the use of generative AI had been previously proposed.”
    • Also stated in: § 2 METHODOLOGY · plain
      “The scarcity of results highlighted the novelty of the concept and the current gap in academic literature.”
    • Also stated in: § 3.4 Need for a Generative AI Literacy competency model · plain
      “the database search showed the absence of literature defining the concept of generative AI literacy and also a comprehensive framework that systematically addresses competency acquisition in navigating and utilizing generative AI tools.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    AI literacy does not equate to computing literacy, nor is computing literacy necessary for AI literacy.

    “As noted, AI literacy does not equate to computing literacy, nor is computing literacy necessary for AI literacy, as individuals may recognize, learn about, and use AI without necessarily being able to program or having experience in computer science and engineering [61].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [61]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Premisereasoning

    Existing AI literacy frameworks tend to be quite generic and fail to address the specificities of generative AI tools.

    “the existing frameworks on AI literacy tend to be quite generic, failing to address the specificities of generative AI tools.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    A number of prompt engineering techniques have emerged, including automatically generated prompts, Chain of Thoughts prompting, and Tree of Thoughts prompting.

    “A number of techniques have emerged for this purpose, including automatically generated prompts [86], Chain of Thoughts prompting [50,98,107] and Tree of Thoughts prompting [62,104].”
    Quote in the paperNot a substantial claim0 of 6 sources checked
    • cites [86]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [50]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [98]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [107]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [62]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [104]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Guidelines and techniques to help users identify which images, videos, and text have been created by AI are already being discussed.

    “Guidelines and techniques to help users identify which images, videos, and text have been created by AI [39] are already being discussed.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [39]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    The proposed competency model goes beyond prompt engineering and the more technical features of generative AI, considering contextual and situational aspects.

    “while prompt engineering is one of the competencies and is crucial for generative AI literacy, the proposed model goes well beyond it and the more technical features of generative AI, and considers the contextual and situational aspects.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The rapid rise of generative AI tools and their widespread adoption in various disciplines have created a unique set of skills specific to these novel tools, which go beyond the traditional competencies associated with data literacy, digital literacy, or AI literacy.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    Fixing hyperparameters is an additional and more advanced skill in mastering prompt engineering.

    “Hence, fixing these parameters is an additional and more advanced skill to know when mastering prompt engineering.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Prompt engineering consists of a broad skill set, with different methods that allow effective interaction with generative models.

    “Prompt engineering consists of a broad skill set, with different methods that allow to effectively interact with generative models [82].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [82]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationnone given

    Generative AI literacy goes beyond technical proficiency and requires combining it with other proficiencies such as adaptive usage, ethical considerations, and legal awareness.

    “Generative AI literacy goes beyond technical proficiency and requires combining it with other proficiencies such as adaptive usage, ethical considerations, and legal awareness.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationnone given

    Generative AI literacy is not a static set of skills, but an evolving one.

    “Generative AI literacy is not a static set of skills, but an evolving one.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Generative AI technology is evolving rapidly, with new models and techniques continually emerging, each with its own set of intricacies.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Individuals in organizations already use or can use generative AI applications for tasks such as note-taking, summarization, time management, and providing faster support to common queries.

    “As part of organizations, individuals already use (or can use) generative AI applications for tasks such as note-taking, summarization, time management, and providing faster support to common queries [12].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [12]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    In 2023, President Joe Biden issued Executive Order 14110 on Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence.

    “in 2023 President Joe Biden issued Executive Order 14110 on Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence [40]”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [40]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Generative AI has the potential to ease access to internal knowledge and databases through chat and provide answers to frequently asked questions by citizens.

    “In addition to its role in supporting policy development and regulation, generative AI has the potential to ease access to internal knowledge and databases through chat and also provide answers to frequently asked questions by citizens [28].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [28]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    The competencies outlined in the paper are not confined to specialists; they cater to a broader audience.

    “Thus, the competencies outlined here are not confined to specialists; rather, they cater to a broader audience, reflecting the growing integration of AI into various aspects of daily life.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    An example of an AI-generated image misleading voters occurred during the campaign for Switzerland's 2023 elections.

    “An example of such a scenario occurred during the campaign for Switzerland's 2023 elections [99].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [99]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    AI is already employed at various levels in the public sector, including service delivery, management, and monitoring.

    “AI is already employed at various levels in the public sector, including service delivery, management, and monitoring [10,43,64,95].”
    Quote in the paperNot a substantial claim0 of 4 sources checked
    • cites [10]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [43]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [64]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [95]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    At the time of writing, only a few countries have begun a legislative process regulating AI and its applications.

    “While at the time of writing this paper, only a few countries have begun a legislative process regulating AI and its applications, it is likely that several others will follow [37,65].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [37]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [65]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    In March 2024, the European Parliament approved the European AI Act, determining a set of norms for several aspects of AI.

    “Indeed, in March 2024 the European Parliament approved the European AI Act [32], determining a set of norms to be followed with respect to several aspects of AI at large [41,42,63].”
    Quote in the paperNot a substantial claim0 of 4 sources checked
    • cites [32]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [41]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [42]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [63]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Several universities have developed guidelines for the use of LLMs in assignments, differing from campus to campus, with some prohibiting LLMs altogether and others allowing them for certain coursework.

    “Several universities have developed guidelines for the use of LLMs in assignments [26].This differs from campus to campus, with some of them prohibiting LLMs altogether, and others allowing them for certain types of coursework.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [26]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Several states from the United States have begun to adopt laws targeting specific aspects of AI.

    “several states from the United States have begun to adopt laws targeting specific aspects of AI [17]”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [17]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

31 of 31

  • The paper relies on this without support.

    The rapid rise of generative AI tools and their widespread adoption in various disciplines have created a unique set of skills specific to these novel tools, which go beyond the traditional competencies associated with data literacy, digital literacy, or AI literacy.

    “the rapid rise of generative AI tools and their widespread adoption in various disciplines have created a unique set of skills specific to these novel tools, which go beyond the traditional competencies associated with data literacy, digital literacy, or AI literacy.”

  • The paper relies on this without support.

    Generative models have the potential to fabricate highly realistic content and give rise to ethical dilemmas about misinformation, authenticity, and societal impact.

    “Generative models are ingrained with the potential to fabricate highly realistic content and give rise to ethical dilemmas about misinformation, authenticity, and societal impact.”

  • The paper relies on this without support.

    The outputs of generative AI raise critical questions on authorship, ownership, and originality.

    “In the third place, the outputs of generative AI raise critical questions on authorship, ownership, and originality.”

  • The paper relies on this without support.

    Generative AI tools are increasingly becoming accessible and user-friendly, necessitating additional aspects of AI literacy for the general public, not just for people specialized in technology.

    “Finally, generative AI tools are increasingly becoming accessible and user-friendly, necessitating additional aspects of AI literacy for the general public, not just for people specialized in technology.”

  • The paper relies on this without support.

    Most prompt engineering techniques include an assessment of the output.

    “most prompt engineering techniques include an assessment of the output, but this does not necessarily imply that the user is fully proficient in prompt engineering.”

  • The paper relies on this without support.

    Most users will neither have nor need programming and fine-tuning ability for their specific goals, and can still be literate on generative AI without it.

    “Most users will neither have nor need this ability for their specific goals, and still be literate on generative AI without it.”

  • The paper relies on this without support.

    Generative AI technology is evolving rapidly, with new models and techniques continually emerging, each with its own set of intricacies.

    “Generative AI technology is evolving rapidly, with new models and techniques continually emerging, each with its own set of intricacies.”
7 of 7

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Tools designed to accelerate research can, when used uncritically, undermine the very skills graduate education is meant to cultivate.

    “The result is a paradox: tools designed to accelerate research can, when used uncritically, undermine the very skills graduate education is meant to cultivate [1].”

    From Students' Practices and Skills in the LLM-Era: "You Can't Outsource the Struggle and Still Get the Skill" (2026) · cites this paper as [1]

    Section 5.6 says users need critical thinking and should verify AI-generated output rather than use it uncritically. The paper does not establish that research-acceleration tools undermine the skills graduate education is meant to cultivate; its discussion of curricula in Section 6.3.3 concerns AI literacy across educational levels, not that specific effect. (LLM, conf=0.83)

    Experimental — accuracy not yet measured

  • The innovation dimension relates to developing and tailoring new AI tools and workflows to specific educational needs and engaging in human-AI co-creation.

    “The innovation dimension relates to developing and tailoring new AI tools and workflows to specific educational needs and engaging in human-AI co-creation (Annapureddy et al., 2025).”

    From An integrated AI literacy framework for higher‑education teachers: a critical review and conceptual synthesis (2026) · cites this paper as Annapureddy et al., 2025

    Section 5.8 describes developing generative AI models and adapting them to specific applications, while §6.3.7 calls for effective human–AI collaboration. The paper does not define an “innovation dimension” encompassing educational workflows or specifically describe human–AI co-creation for educational needs; §6.3.3 discusses education in the different context of AI-literacy curricula. Page numbers are unavailable in our copy. (LLM, conf=0.85)

    Experimental — accuracy not yet measured

  • A few studies have proposed AI literacy frameworks for higher education, identifying core competencies such as prompt engineering, ethical awareness, and critical evaluation of AI outputs.

    “A few studies have proposed AI literacy frameworks for higher education, identifying core competencies such as prompt engineering, ethical awareness, and critical evaluation of AI outputs [1, 22].”

    From Students' Practices and Skills in the LLM-Era: "You Can't Outsource the Struggle and Still Get the Skill" (2026) · cites this paper as [1]

  • The synthesis identified a significant gap in higher education faculty's innovative use of AI, including designing or customising AI technologies.

    “The study identified significant gaps in ethical understanding, innovative use of AI and institutional support among faculty.”

    From An integrated AI literacy framework for higher‑education teachers: a critical review and conceptual synthesis (2026) · cites this paper as Annapureddy et al., 2025

    Section 5.10 discusses ethical understanding as a generative AI literacy competency, and Section 6.3.3 discusses curricula for students. Neither reports gaps among higher education faculty in innovative AI use, ethical understanding, or institutional support. Searches for faculty gaps and designing or customising AI technologies found no passage addressing that finding; page numbers were unavailable in our copy. (LLM, conf=0.72)

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Pentagon example changes from a video to a photo

    Minor

    The example describes a viral AI-generated video, whereas both cited references describe the incident as a fake picture or photo.

  • AI-literacy definition is cited to the wrong reference

    Minor

    The text attributes the definition to Long and Magerko as reference [61], but reference 61 is a paper by Leo S. Lo. Long and Magerko are listed as reference 62.

  • NIH competency definition is cited to the wrong reference

    Minor

    The quoted competency definition is attributed to the National Institute of Health as reference [70], but reference 70 is by Noh and Han. The NIH entry is reference 71.

3 of 3

Cites

Flagged first, then by citations

Research works printed but not linked: 2

  1. Yoon Chae. 2020. US AI regulation guide: Legislative overview and practical considerations. Journal of Robotics, Artificial Intelligence & Law 3 (2020), 17–40.
  2. Tambiama Madiega. 2021. Artificial Intelligence Act. European Parliamentary Research Service.https://policycommons.net/artifacts/1894362/artificial-intelligence-act/2644377/#::text=https%3A//policycommons.net/artifacts/1894362/artificial%2Dintelligence%2Dact/2644377/

Cited by

Sorted by citations

Related

Authors
Annapureddy, R., Fornaroli, A., & Gatica-Perez, D.
Affiliations
  • École Polytechnique Fédérale de Lausanne (CH)
  • Idiap Research Institute (CH)
  • Idiap Research Institute, Martigny, Switzerland and EPFL, Lausanne, Switzerland
ORCID
Venue
Digital Government: Research and Practice
Publisher
Association for Computing Machinery (ACM)
Published
2025-02-13
Volume
6
Issue
1
Pages
1-21
ISSN
  • 2639-0175 (ISSN-L)
  • 2691-199X
DOI
10.1145/3685680
OpenAlex
W4401288535
Cited by
333
References
95
Citations by year
2026: 245 · 2025: 81 · 2024: 7
FWCI
166.06
Type
journal article
Language
en
Open access
Licence
www.acm.org/publications/policies/copyright_policy#Background (vor)
Locations
  • Digital Government Research and Practice (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • Infoscience (Ecole Polytechnique Fédérale de Lausanne) (submittedVersion, cc-by-nc-nd)
  • arXiv (Cornell University)
Funders
  • European Union’s Horizon 2020 IcARUS project (882749)
  • Horizon Europe ELIAS project (101120237)
Topics
  • Physical Sciences › Computer Science › Information Systems › Digital literacy in education
  • Social Sciences › Safety Research › Ethics and Social Impacts of AI
  • Physical Sciences › Computer Science › Information Systems › Artificial Intelligence in Education
Keywords
generative AI literacy, competency-based model, prompt engineering, AI ethics, responsible AI use, artificial intelligence in education, professional training, AI regulation

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-11-06.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-11-06.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review