Theoretical Framework for AI-Integrated Project 

This project is grounded in three key pedagogical and technological models that justify the use of Adobe Firefly for a B1-level, research-based Instagram campaign. These frameworks establish the project as an advanced form of digital learning designed to cultivate high-order thinking skills appropriate for the AI era.

1. SAMR Model (Chouduri, 2023)

According to Romrell et. al. (2014), the SAMR model, developed by Ruben Puentedura, is a framework used to classify how technology is integrated into learning. It organizes technology use into four levels, moving from simple enhancement to complete transformation.It goes from a very basic use to a more elaborate one, in which the whole focus is changed. The first level, Substitution, involves using technology to replace a traditional tool without any change in function. For example, typing a document on a computer instead of writing it by hand. The next level, Augmentation, uses technology to improve a task with functional enhancements, such as using a word processor with a spell-checker. The third level, Modification, is a more significant step where technology allows the learning task to be redesigned. The final and most transformative level, Redefinition, involves creating entirely new learning activities that were previously impossible without the use of technology.

Choudhuri (2023) adapts the SAMR model to analyze the integration of AI in education, moving from simple replacement to transformative redefinition. The author explains that Substitution occurs when AI automates existing tasks, like generating a worksheet, which saves time but doesn't change core teaching practices. Augmentation enhances content delivery, while Modification begins to reshape teaching methods. The ultimate goal, however, is Redefinition, where AI enables entirely new learning experiences that were previously impossible, such as personalized learning. The author argues that while most current AI tools for educators are stuck at the substitution and augmentation levels, the true potential of AI lies in its ability to redefine education and amplify human capabilities, creating innovative and personalized learning opportunities that address long-standing educational challenges.

The SAMR model (Substitution, Augmentation, Modification, Redefinition) is used to analyze and justify the level of technology integration in the class. The model describes how technology adoption moves from simple replacement to functional transformation. Substitution uses technology as a direct tool replacement with no functional change; Augmentation uses technology as a replacement with functional improvement; Modification allows for significant task redesign; and Redefinition enables the creation of new tasks previously inconceivable. This project  targets Modification (M) and achieves Redefinition (R). The use of Adobe Firefly significantly redesigns the concluding project phase: instead of a traditional poster, students use AI to quickly iterate complex visual concepts (Modification). Furthermore, the final output—an AI-generated image combined with a B1-level caption for an actual public-facing Instagram awareness campaign—creates a new task previously inconceivable, moving the project to Redefinition by providing a real-world audience and consequence.As Chouduri (2023) expresses levels of Modification and Redefinition show that technology is not just facilitating learning but is fundamentally transforming the learning process and its results.In this case, the way in which technology helps to make the message spread is also a fundamental change not only of how we access information but in hat we do with it.

2. Connectivism (Bates, 2023)

Coonectivism is a framwork that attempts to account for the way people learn now: it is a play on words becuase it combines cognitivism and connection, because now we learn with others, through technology. In this vision, we learn by connecting ideas via technology, by producing as well, not simply by consuming information. there is a special focus on how people share information, too.We no longer learn alone, but in sharing information with others.

Connectivism emphasizes learning as a process: we connect sources and form networks. From this perspective, we know that sharing with others is crucial, it is essential for justifying the peer feedback and public campaign aspects of the plan. Students access and utilize diverse nodes in their learning network: their prior research data (internal), the AI tool (Firefly) (external), and the peer feedback network (Classroom) (social). The task prioritizes the skills of curating, managing, and navigating this digital information (research findings, visual prompts, social media response). By launching the campaign on Instagram and later reviewing analytics data (Beyond stage), students engage in learning that occurs out of the classroom and is driven by current, changing information about public engagement. It also connects the learning experience, students will have studied online gambling, with their context and then revising the impact of their work in their community. In this way, the work is more relevant and motivating for students.

As Bates explains, in a digital world, the challenge is not simply storing knowledge, but knowing where to find, evaluate, and combine the knowledge needed for effective action.

3. AI-Revised Bloom's Taxonomy (Jain & Samuel, 2025)

This interesting framework has analyzed Bloom's taxonomy to account for the used of AI, in education. It has also restructured other elements becuase it is no longer hierachical, elements combine differently. The new taxonomy challanges the idea of low order thinking and high order. This is no longer considered this way. The new taxonomy includes ventriloquising, cocurating, and critical thinking at a later stage. Human beings are still at the centre but they are closely assisted by technology.

This framework provides the specific cognitive structure, validating the objectives that move beyond traditional high-order thinking to account for co-piloted learning, where the AI acts as a partner. The project is structured around the three key AI-integrated skills. Critical Understanding requires students to engage in deep analysis of their research findings to select the most impactful visual message, then evaluate and critique the ethical and emotional efficacy of the AI-generated outputs, going beyond simple analysis by requiring ethical and audience justification. Ventriloquising is the core skill of directing the AI; students must use precise B1 English (the "voice") to instruct the AI to visualize a concept that is true to their research, replacing simple application.As Jain & Samuel(2025)  explain, Ventriloquising is to master  the textual prompt to instruct the AI, transforming the user from a passive receiver into an active, intentional director of creative output. explain Finally, Co-curating redefines the traditional Creating stage, as the final task is not just generating an image, but combining the AI output (the image) with the human-authored text (the B1 caption and hashtags) to create a final, meaningful, and coherent campaign asset. 

Interestingly, if students work on the topic suggested of online gambling, using technology, they will find what a threat this may represent for chidren their age. By sharing their findings and reflecting on the results from their Instragram campaing, the critical thinking stage is reinforced.