The Paradox of AI in Educational Frameworks: Enhancing Student Autonomy While Risking Dependency

The integration of artificial intelligence (AI) into educational frameworks such as SCORE (often aligned with self-regulated learning, SRL) is profoundly reshaping how students govern their educational trajectories. Recent systematic reviews and empirical studies indicate that AI tools—ranging from conversational agents to adaptive feedback platforms and generative models—can facilitate the core phases of SRL: forethought (planning), performance (strategy utilization), and reflection (self-assessment). Consequentially, these technologies hold significant promise in supporting student autonomy and intrinsic motivation.

However, these benefits are heavily nuanced by growing concerns regarding over-reliance on automated systems. Extensive deployment without adequate pedagogical scaffolding may inadvertently reduce learners' initiative, blunt critical thinking, and dampen self-efficacy. Current literature exposes a critical paradox: while AI empowers students to assume greater control over their learning vis-à-vis personalization and immediate feedback, it also cultivates vulnerability to dependency if learners outsource substantial cognitive effort to the technology. This review synthesizes findings from recent high-impact research to elucidate how AI-SCORE integration alters the delicate balance between student autonomy and dependency across diverse educational landscapes.

A rigorous and comprehensive systematic search was conducted across multiple major scientific databases, including Semantic Scholar and PubMed. The initial retrieval yielded a vast corpus of potential records. After applying relevance filtering and removing duplicates, 161 manuscripts were retained for eligibility screening based on predefined inclusion criteria. Ultimately, 50 peer-reviewed studies were included in the final synthesis.

Six unique search strategies were employed targeting: (1) SCORE/SRL frameworks, (2) AI integration effects on autonomy versus dependency, (3) adjacent theoretical models (e.g., self-determination theory), (4) critical evaluations of technological over-reliance, (5) foundational works on learner agency, and (6) experimental findings on AI-driven metacognition.

**Positive Impacts: Fostering Autonomy & Motivation** AI tools augment SRL by structurally enhancing goal-setting, strategic planning, task engagement, and reflective practices. Adaptive systems provide personalized feedback that assists students in intuitively monitoring their academic progress and modulating their learning strategies. Empirical evidence establishes that when thoughtfully integrated—particularly under active teacher mediation—AI bolsters digital efficacy, intrinsic motivation, and a sustained willingness for autonomous learning.

**Risks: Dependency & Over-Reliance** Conversely, excessive reliance on AI for planning or automated feedback undermines learner autonomy by offloading necessary cognitive exertion. Over-dependence reliably diminishes a student's confidence in independent problem-solving and engenders "metacognitive laziness," a state wherein students passively defer to automated suggestions rather than critically engaging with the academic material.

**Nuanced Outcomes: Ambivalence & Contextual Factors** Learners frequently report epistemic ambivalence—simultaneously operating as independent agents while leaning heavily on AI for efficiency and assessment tasks. Contextual variables are highly disruptive: higher education students generally extract more autonomy-related benefits from AI than younger cohorts, while direct teacher involvement serves as a potent moderator, often increasing motivation while occasionally compressing perceived autonomous space. Furthermore, affective states, such as positive emotional valence during AI usage, significantly mediate a learner's willingness to engage autonomously.

**Design & Implementation Considerations** Effective integration demands a precise equilibrium between providing technological scaffolding and commanding self-reflection and critical engagement. Intentional pedagogical strategies—such as explicit instruction on responsible AI utilization and the cultivation of advanced digital literacy—are mandatory to forestall superficial application. Human-centered agency, anchored by active teacher involvement, remains the critical bulwark against systemic AI dependency.

The aggregated research forcefully argues that integrating AI within the SCORE/SRL framework offers profound capabilities to elevate student autonomy via hyper-personalized pathways and instantaneous, actionable feedback. Nonetheless, this technological empowerment is highly conditional, heavily reliant on sophisticated pedagogical design. In the absence of deliberate safeguards against passive consumption, AI poses a tangible threat: the erosion of self-efficacy and the atrophy of critical thinking skills.

The dualistic nature of these outcomes underlines the absolute necessity for instructional methodologies that explicitly promote metacognitive awareness in tandem with technological support. While high-quality evidence—derived predominantly from higher education experiments and systematic reviews—substantiates these claims, critical voids persist regarding longitudinal durability and cross-demographic variances. The threat of cognitive outsourcing necessitates a paradigm shift from viewing AI merely as an assistive tool to treating it as a complex cognitive partner that requires rigorous governance.

The matrix below shows where empirical evidence is concentrated and where critical research gaps remain.