Self-Determination Theory in the Age of Generative AI: Restoring Autonomy, Competence, and Relatedness in Higher Education Assessment

In 1985, psychologists Edward L. Deci and Richard M. Ryan introduced Self-Determination Theory. Over four decades of research across universities, they demonstrated that human motivation operates from external compliance to intrinsic engagement. Deep learning requires satisfying three basic psychological needs: autonomy, competence, and relatedness. When students experience volition, master challenges, and connect with faculty and peers, conceptual understanding develops naturally.

Generative artificial intelligence introduces a serious motivational challenge in higher education. Large language models provide instant prose, code, and summaries. When learners bypass the friction of drafting, revising, and problem solving through prompts, the link between effort and outcome dissolves. Students submit deliverables reflecting machine capability rather than mastery.

Many universities responded with surveillance, deploying unreliable statistical detectors and remote proctoring. These measures treat students with suspicion, enforce compliance through fear, and extinguish autonomy. As documented in the collapse of the artifact economy, grading static text submissions no longer verifies understanding. Resolving this challenge requires redesigning assessment around human motivation.

The Three Pillars of Self-Determination Theory

Understanding how generative tools disrupt learning requires examining three core psychological needs identified by Deci and Ryan:

1. Autonomy

Autonomy refers to the experience of volition and agency. Autonomous learners perceive themselves as authors of their actions, engaging in work because they value knowledge or professional relevance. Autonomy cannot coexist with coercion. When learners possess meaningful choices regarding how they investigate problems, commitment to integrity deepens.

2. Competence

Competence represents the feeling of effectiveness in one's environment. Competence grows out of productive struggle: encountering difficult concepts, testing hypotheses, and achieving clarity through effort. Real mastery requires overcoming intellectual difficulty.

3. Relatedness

Relatedness represents the need to feel connected and valued within a community. In university courses, relatedness grows through intellectual dialogue, collaborative inquiry, and mentorship. Students thrive when instructors treat them as developing scholars rather than anonymous records in a gradebook.

The Motivational Crisis: How Generative AI Undermines Basic Needs

Conventional grading of static written products destabilizes all three pillars when generative AI is introduced:

First, generative AI erodes competence. When an undergraduate uses automated assistants to draft an essay, the student knows they did not perform the underlying synthesis. This produces an imposter dynamic: synthetic fluency masking internal self-doubt. The student never experiences the satisfaction of overcoming intellectual difficulty, causing self-efficacy to decline.

Second, institutional surveillance destroys autonomy. Policing generative AI through automated detectors or camera monitoring turns assessment into an adversarial contest. Students feel distrusted. Extrinsic pressures crowd out intrinsic motivation, reducing education to compliance. As detailed in our analysis of alternatives to AI detection, algorithmic surveillance punishes students while failing to confirm understanding.

Third, solitary interaction with chatbots degrades relatedness. Replacing peer discussions and office hours with isolated prompting separates learners from scholarly communities. Higher education becomes transactional: students prompt software, submit files through learning management systems, and receive automated scores. Intellectual community vanishes, leaving students isolated.

Shifting from Extrinsic Compliance to Autonomous Motivation

Self-Determination Theory organizes human motivation along an explicit continuum:

• External Regulation: Action driven purely by tangible rewards or punishments, such as avoiding a failing grade.
• Introjected Regulation: Action driven by internal guilt or anxiety about academic status.
• Identified Regulation: Action driven by conscious valuing of knowledge, such as an engineering student mastering thermodynamics.
• Integrated Regulation: Action harmonized with personal identity, ethics, and professional goals.
• Intrinsic Motivation: Action undertaken out of genuine curiosity, challenge, and subject interest.

Traditional assessments frequently trapped students in external and introjected regulation. When faculty assigned term papers or exams, students focused on points and penalties. Generative AI makes external compliance trivial to automate: software fulfills rubric checklists without engaging student thought.

To cultivate identified and intrinsic motivation, assessment must shift from grading passive written products to evaluating active reasoning. When faculty ask students to articulate choices, defend methodology, and analyze real-world constraints, students regain an authentic stake in their work.

Spoken Defense as an Engine of Competence and Agency

Structured oral defense provides an effective mechanism to satisfy competence. Students submit their project, code repository, or paper, then verbally explain their reasoning during a brief dialogue. Guided by faculty questions, students justify conclusions, trade-offs, and responses to hypothetical alterations in source material.

This workflow transforms the psychology of assessment:

• Overcoming Synthetic Imposter Syndrome: Explaining ideas aloud confirms understanding. Speaking creates clear evidence of competence that chatbots cannot simulate. The student realizes that while AI may assist in preliminary brainstorming, the core reasoning belongs to them, replacing self-doubt with authentic self-efficacy.
• Cultivating Voice and Authorship: Spoken defense restores autonomy by centering student voice. Students take ownership of their arguments as emerging scholars, gaining genuine agency over scholarly identity.
• Providing Nuanced Diagnostic Feedback: As established in our research on formative assessment in the age of generative AI, spoken dialogue reveals the boundary between solid mastery and confusion. Instructors validate student insight directly, guiding future study with precision.

Interactive Simulations: Designing Autonomy-Supportive Problem Spaces

Autonomy thrives when students navigate environments requiring independent decisions. Interactive scenario simulations provide realistic problem spaces where students apply theory to dynamic challenges.

Rather than reading static case studies and typing easily automated answers, students navigate simulations where choices trigger realistic consequences:

• Healthcare and Clinical Practice: Nursing students manage an evolving patient scenario, prioritizing clinical interventions as vital signs fluctuate, as discussed in our examination of diagnostic assessment in the age of generative AI.
• Business and Strategic Management: Marketing students negotiate terms with AI-driven stakeholder personas, balancing corporate priorities and budget limits while justifying concessions.
• Environmental Science and Public Policy: Students balance economic development against environmental preservation during a municipal crisis, adapting recommendations as community stakeholders raise objections.

In each scenario, students exercise genuine autonomy without scripted paths. Students assess incomplete information, select actions, observe outcomes, and reflect on results. This iterative process satisfies the need for competence while preparing students for professional practice, demonstrating how how to turn a case study into an interactive AI simulation elevates course design.

Restoring Relatedness Through Dialogic Assessment

Authentic assessment repairs the relational fabric of higher education. When evaluation shifts toward dialogue, the relationship between instructor and student transforms.

Under surveillance models, faculty serve as examiners searching for infractions. In dialogic assessment, instructors act as intellectual mentors. Reviewing spoken defenses or simulation debriefs centers on disciplinary inquiry: exploring why a student chose a specific model, or how an intervention would change under altered constraints.

This interaction validates the student as a contributing scholar. As explored in our guide to evaluative judgment in higher education, dialogic interactions teach students how professional communities debate evidence and reach consensus. Relatedness becomes the social anchor motivating persistence through challenging hurdles.

Sustainable Implementation and Faculty Workload

Faculty often worry that oral assessments and simulations will overwhelm grading schedules. In a lecture hall of 120 undergraduates, conducting face-to-face examinations during office hours is impractical.

Institutions maintain high relational quality and competence verification without excessive grading burdens:

• Asynchronous Spoken Submissions: Students record concise two-minute spoken defenses through their browser or LMS, responding to tailored prompts probing their submitted work without scheduling conflicts.
• Synchronized Audio and Transcript Review: Instructors review recordings alongside synchronized verbatim transcripts, allowing rapid scanning and targeted playback in two to three minutes per submission.
• Rubric-Aligned Scoring Frameworks: Structured criteria focus on conceptual grasp, methodological justification, and problem solving. As detailed in our guide on how oral assessment works in large classes, asynchronous workflows scale to large cohorts without extending grading hours.

Universal Design and Accessibility in Autonomy-Supportive Assessment

Supporting student autonomy requires assessments that respect learner variability. Forcing students into high-pressure synchronous interrogations induces performance anxiety, skewing evaluation away from academic competence toward social confidence.

Autonomy-supportive assessment incorporates Universal Design for Learning (UDL) principles, as examined in universal design for learning in the age of generative AI:

• Multiple Means of Action and Expression: Offering asynchronous video, audio-only recordings, or timed interactive transcripts ensures students with speech challenges or anxiety demonstrate competence effectively.
• Formative Practice Rehearsals: Practicing in low-stakes simulation environments builds familiarity and reduces evaluative threat. As established in metacognition in the age of generative AI, structured self-reflection helps students monitor preparation.
• Objective, Bias-Mitigated Rubrics: Defining evaluation criteria ensures grading reflects disciplinary understanding rather than accent, dialect, or vocal cadence, as detailed in our analysis of grading bias in oral assessment.

How Rocketproof Rebuilds Student Motivation and Agency

Rocketproof is designed around the insight of Self-Determination Theory: students learn best when they experience autonomy, build authentic competence, and engage in meaningful dialogue.

Rocketproof provides faculty with an authentic-assessment platform evaluating the student rather than the submission alone. Instead of attempting to detect AI writing through unreliable probability scores, Rocketproof helps instructors create structured oral defenses, interactive simulations, and source-grounded discussions:

• Source-Grounded Questioning: Rocketproof analyzes submitted projects, papers, or code, generating targeted probing questions testing whether students understand their citations, data structures, and conclusions.
• Scalable Spoken Defense: Students record short spoken responses within Canvas, Blackboard, Brightspace, or Moodle. Instructors receive synchronized audio-transcript interfaces with rubric-aligned evaluations, making oral assessment feasible for any cohort size.
• Interactive Disciplinary Simulations: Faculty configure branching role-play and decision-making scenarios without writing code, allowing students to navigate professional challenges with autonomy.
• Faculty Control: Rocketproof never automates grading decisions. Instructors maintain complete authority over scores, rubrics, and feedback, preserving the mentor-student relationship.

Explore our oral defense platform, discover authentic assessment, review our interactive AI simulations, or examine our LMS integration. To discuss a pilot for your institution, contact our academic team.

Frequently Asked Questions

What is Self-Determination Theory, and why is it relevant to generative AI in higher education?

Formulated by Edward L. Deci and Richard M. Ryan, Self-Determination Theory states that human motivation and learning depend on satisfying three basic psychological needs: autonomy, competence, and relatedness. Generative AI disrupts this balance by automating cognitive tasks, bypassing the productive struggle required for competence while triggering surveillance that crushes autonomy.

How does generative AI undermine student feelings of competence?

When students use generative AI to produce assignments, they receive credit for work they did not intellectually author. This produces an illusion of competence, eroding self-efficacy and creating imposter anxiety because students realize they cannot replicate the quality without algorithmic assistance.

Why does institutional surveillance fail to restore academic integrity?

Adversarial measures such as automated AI detection and remote monitoring treat students with suspicion. These controls undermine autonomy, heighten anxiety, and push learners into compliance, encouraging evasive behaviors rather than intrinsic motivation.

How does a spoken defense rebuild student autonomy and agency?

A spoken defense requires students to verbally articulate decisions, explain trade-offs, and defend conclusions. Speaking ideas aloud transforms students from passive submitters of synthetic text into active scholars who take public ownership of their intellectual contributions.

How can faculty implement spoken assessments without increasing grading workloads?

Instructors use asynchronous oral assessment tools integrated into the learning management system. Students record ninety-second responses to analytical questions, which instructors evaluate using synchronized audio-transcript dashboards in under three minutes per student.

Reclaiming the Human Purpose of Higher Education

The challenge generative artificial intelligence poses to higher education is psychological and pedagogical rather than merely technological. When universities treat assessment as an administrative audit of written deliverables, students adopt automated shortcuts fulfilling external requirements with minimal engagement.

Self-Determination Theory provides a clear path for institutional renewal. Higher education achieves its purpose when nurturing self-determined, competent, and socially connected thinkers. By replacing passive artifact grading and surveillance with spoken defense, interactive simulations, and dialogic inquiry, universities create environments where students experience genuine mastery.

When institutions assess the student rather than the submission, they restore the joy of learning, verify authentic capability, and prepare graduates for purposeful professional lives.