As artificial intelligence (AI) has begun to integrate into all aspects of daily life, the challenge of better preparing teachers’ education to employ intelligent technologies effectively and efficiently in schools has become a persistent issue ^[1][2][1,2]. Artificial intelligence has become the core technology in education, but it has no straightforward narrative from the 1950s until now. Progressive advancements in AI have generated intelligent machines and algorithms that can know and adapt in response to the environment and sets of rules which simulate human intelligence. HeAccoreinding to this study, AI refers to the science and technology of research and development of theories, methods, techniques, and application systems for simulating and extending human intelligence, which is connected to the challenge of utilizing computers to comprehend or imitate the intelligence of humans without any logical order or algorithm ^[3][4][3,4].

Traditionally, isolated ICT courses or units have been provided to promote teacher education, often supplied in early teachers’ improvement programs. These are conducted with the assumption that the strategy of “front loading” teachers with what is considered the necessary knowledge and skills to transform the information [5] will support them in meeting course assessment requirements, such as the growth of “technology-integrated” learning curricula for practical work in schools, and help them use AI technology productively in their future teaching careers ^[6][7][6,7]. Artificial intelligence technology has become an important driver for educational teaching reform and it affects teachers’ sustainable professional development. However, AI education courses and training opportunities are more available to students than teachers. Thus, to extend our understanding of AI literacy for developing the capabilities of teachers, the construction of AI literacy merits theour special attention.

Similar to information literacy and digital literacy, AI literacy is a combination of “technology” and “literacy”, i.e., “artificial intelligence” and “literacy”, and is used to define skill sets across disciplines [8]. According to the study of Ng et al., AI literacy can be divided into four common dimensions to build structure: (1) the understanding level of fundamental AI concepts to support basic AI training; (2) the application of AI concepts in practical backgrounds to facilitate widespread AI education; (3) the evaluation and engagement with AI technologies within those contexts should be critical and reasonable; and (4) the ability to understand the endless ethical implications resulting from AI applications [9]. Based on the dimensions that already have been defined, there wasis study analyzed the situation of teachers’ AI literacy. The AI literacy of teachers should be distinct from practical contexts with stricter requirements since these dimensions have not yet been targeted and tested in the course of teacher training [10].

Literacy was universally perceived as an ability that arises from learning skills including reading, speaking, and writing [11]. The appearance of knowledge based on AI technology education means that every learner should have the basic competencies of “intelligent literacy” to build a better foundation for integration into the digital society, so as to gain equity and respect in working life. This term has been extended to new areas of literacy, such as computer, information, digital, and AI literacy [12]. Information literacy can be taken as an example which empowers people in all walks of life to seek, evaluate, use, and create information effectively in order to achieve their personal goals, and mainly includes the skills of information management and the ability to apply information in an appropriate way ^[13][14][13,14]. In the twenty-first century, teachers with these technical skills can use advanced intelligent technology through computers to impart new knowledge and skills to students or colleagues in certain avant-garde ways ^[15][16][17][15,16,17].

Currently, with the advent of AI technologies, the application of AI has become critical and plays key roles across disciplines and industries ^[9][18][9,18]. Teachers need to know how to use AI technologies explicitly in their teaching, as well as to take ethical issues into consideration ^[19][20][19,20]. Thus, combining AI and literacy means that AI literacy is about the critical skills that individuals need to learn and live in the digital world with the help of AI-based technologies. According to Bloom, the educational components include the cognitive, psychomotor, and emotional domains, as well as the division of education in terms of knowledge, skills, and attitudes ^[21][22][21,22]. While AI literacy is not equal to the applications of AI, following the conclusion of Bloom, it includes four dimensions: Knowing and Understanding AI, Applying AI, Evaluating AI Application and AI Ethics, as identified from previous studies ^[23][24][25][23,24,25].

As teachers, the most significant AI literacy in education is about fundamental concepts, knowledge, information, and attitudes that require knowledge about AI. As the users of AI applications, teachers should prioritize learning the theoretical knowledge of AI instead of applying AI approaches at random and without prior intention. According to previous studies, AI literacy was initially defined as the ability to know the basic techniques and concepts of artificial intelligence embedded in different products and services ^[26][27][26,27]. A large number of teachers have been exposed to the technological environment with AI-enabled appliances, but in reality, they do not understand the fundamental concepts, such as big data structures, computational thinking concepts, or ubiquitous computing approaches. Sound knowledge about AI and principles of machine learning is so essential for their later careers that teachers should apply AI every classroom. Moreover, some researchers link AI literacy to learners’ self-perception, confidence, and readiness to learn AI. Meanwhile, a poor belief state in the potential of AI among teachers has also been proven ^[28][29][28,29]. To foster the AI literacy of learners, several scholars designed learning curricula and activities that promote AI literacy, paying attention to the way that teachers learned AI concepts ^[23][30][23,30], as well as understand AI. Thus, Knowing and Understanding AI (KUAI) is fundamental AI literacy that teachers should understand the basic concepts, knowledge, and instructions about using artificial intelligence in teaching.

It is obvious that educating teachers to know how to apply AI concepts and applications in different contexts is of great importance ^[23][30][23,30]. Machine learning applications have been evaluated after learning related curricula, which aims to educate citizens to understand AI applications and learn to adapt in their later careers, as well as understanding the ethical issues related to AI technologies ^[31][32][31,32]. Several studies have discussed the human-centered considerations and concentrated on using AI concepts and applications both regularly and ethically ^[33][34][33,34], which will be further discussed in H1. It is a struggle for teachers to guide students with the help of AI applications, since they still lack the ability to efficiently analyze data or integrate AI into instructional design ^[35][36][35,36]. AI thinking refers to building logic and algorithms to support teachers in understanding how to employ their knowledge in order to solve problems, work with unstructured data, and handle semantics issues in a scheduled way [37]. Taking How and Hung’s research into account, data analysis through computing allows teachers to make new discoveries from hidden patterns in the data through machine learning, which is a practical application of AI-based thinking [38]. For a teacher, applying AI is a literacy that educators utilize in terms of when and how to use AI in class exactly and correctly.

Compared with applying AI, the ability to evaluate AI application appropriately is more challenging to a teacher. In addition to understanding and using AI concepts in practice, AI literacy has been extended to two other competencies that enable teachers to evaluate AI technologies correctly and critically and to communicate, as well as collaborate effectively, with AI [8]. For example, enhancing teachers’ scientific and technological knowledge through evidence-based learning and continuous curriculum is crucial, since the aim of learning basic knowledge is to implement it in practice [37]. Individuals in the co-creation of AI facilities in public spaces could expand their literacy and experiences of public AI [39]. Teachers who are able to evaluate AI applications appropriately could inform, support, manipulate, and categorize AI concepts together in novel ways. After all, previous studies suggested slight variations on the definition of AI literacy, which support the idea that everyone should acquire basic AI knowledge and competencies, especially school teachers. Evaluating AI applications could be beneficial for enhancing their motivation and interest in teaching with AI [40]. Apart from the ethics of understanding and using AI, evaluating artificial intelligence applications is an important competency that enables teachers to judiciously evaluate AI technologies, commenting on the application of AI in teaching [41].

As to developing AI literacy, understanding AI concepts plays an essential role for learners to evaluate AI applications, since they had to invoke relevant concepts when interacting with AI services to ensure that the AI evaluation was in line with the educational technology. The content of basic AI concepts is uncertain while it has been discussed in several studies ^[12][38][42][12,38,42]. Deep learning, block chains, and machine learning are common AI concepts explored in the literature due to their broad applicability and huge influence. Technologies, such as machine learning, deep learning, and neural networks are listed as AI concepts required for AI literacy, and in addition to these, technologies such as blockchain and cloud computing are also important extensions of AI concepts ^[43][44][43,44]. Based on learners’ acceptance level, a few studies also conducted AI literacy courses for teachers, with topics such as machine learning and neural networks providing more advanced topics for later sections, stepping up the difficulty of the course to suit the learner’s knowledge development pattern ^[28][45][28,45].

In view of the complexity of AI concepts, support for learning artifacts is important to facilitate the understanding of AI evaluation and increase motivation and interest in learning AI knowledge. Recently, an improvement has emerged in hardware and software of computers and robots that enhance AI concepts available to educators [46]. Considering that Knowing and Understanding AI makes teachers learn more about how to apply AI properly, Evaluating AI Application is also significant based on teachers’ cognition of AI.

Besides emphasizing conceptual and literacy development for teachers, the AI literacy course has an important purpose: to equip participants with the basic skills to apply AI flexibly so that they can confidently engage in the digital world and integrate AI into their daily teaching. Through an understanding of how AI works, teachers will be able to think about how the use of AI can be integrated into education for the development of students and share these ideas with other peers. With initial mastery of AI, teachers will grow in confidence regarding AI use, naturally engage in critical discussions related to AI, and use AI to drive instructional change, thereby experiencing a degree of personal empowerment in the form of increased control over their own lives and coping skills [47]. In AI literacy courses, teachers often work to increase their confidence through exposure to AI, using AI concepts as a foundation and AI assessment to guide use, thus leading to a deeper and more thorough understanding of AI and clarifying the ethical aspects of AI educational applications [48]. Therefore, Applying AI (AAI) is essential to other aspects of AI literacy, directly influencing Knowing and Understanding AI (KUAI), Evaluating AI Application (EAIA) and AI Ethics (AIE).

Given that AI plays a significant role in everyday decision making, it may cause irreparable damage to society if it is not used properly [49]. Scientists and engineers, such as Elon Musk, have a more negative attitude toward AI, with their articulation of the horrors that could be caused by future AI technologies that would pose a serious threat to human existence within decades [50]. However, only a few studies mention the need to consider the concept of human-centered development and drew attention to educating citizens to become socially responsible [51]. Few studies focus on the need for the ethical AI paradox, which found that teachers paid no attention to ethical concerns, such as biased decisions, lack of transparency, privacy issues, and the risk of invasion of privacy ^[52][53][52,53]. Therefore, educators need to focus not only on improving their own AI application skills and interests but also on the potential social implications and ethical issues of AI. To prepare the next generation of responsible citizens for the future, teachers should compete for the use of AI in a reliable, credible, and equitable manner; prioritizing issues of equity, accountability, transparency, and ethics; and promoting the integration of computational technologies into teaching and learning in innovative and responsible ways, while drawing on areas with a sociotechnical orientation to reinvent the shape of education ^[54][55][54,55].

Artificial Intelligence literacy in teacher education is in its infancy. It emphasizes and regulates how teachers should understand AI concepts and apply AI. To expand the use of AI scenarios, there is a need to ensure that AI technologies are designed and utilized in a way that is both inclusive and equitable to addressing the underrepresentation of people of color and women in AI-using populations. Thus, the relationship among AI Ethics, Knowing and Understanding AI, and Evaluating AI Application should be found to affect teachers’ AI literacy.