AI4CAREER: Responsible AI for STEM Career Development at Scale in K-16 Education

Based on the provided text, here is a detailed technical summary of the workshop proposal "AI4CAREER: Responsible AI for STEM Career Development at Scale in K-16 Education."

1. Problem Statement

The rapid integration of Artificial Intelligence (AI) into K–16 education infrastructure is fundamentally altering how students explore, assess, and prepare for STEM careers. While AI offers opportunities for personalized feedback and expanded access to opportunity information, its current deployment presents significant risks:

• Lack of Developmental Alignment: AI systems are often introduced without scaffolding for critical engagement or alignment with specific developmental stages (from elementary inquiry to postsecondary advising).
• Reproduction of Inequities: There is a high risk that AI-driven recommendation and predictive systems will amplify existing structural disparities, bias, and fragmented implementation across diverse districts and institutions.
• Premature Narrowing of Pathways: Predictive labeling based on early performance signals may prematurely constrain students' perceived possibilities, narrowing their STEM identity and career trajectories.
• Governance Gaps: There is no consensus on the appropriate boundaries for AI in career decision-making, how to define "career readiness" in an AI-mediated context, or how to govern these systems to ensure equity and student agency.

2. Methodology

This paper does not present empirical experimental results but rather proposes a structured, interactive half-day workshop designed to generate consensus and frameworks. The methodology for the workshop includes:

• Target Audience: A curated group of 25–35 participants (researchers, educators, practitioners, policymakers) from disciplines including Learning@Scale, AI in Education, Human-Computer Interaction (HCI), and Educational Policy.
• Submission Model: Participants submit 2–4 page position papers or work-in-progress papers to serve as the basis for discussion.
• Workshop Structure (3.5 Hours):
  • Lightning Talks: Ten 4-minute presentations to frame key tensions (definitions, developmental differences, equity, ethics).
  • Activity 1 (Developmental Lens): Small groups analyze AI roles and risks across specific age stages (Middle School, High School, Undergraduate).
  • Activity 2 (Stakeholder Lens): Groups reorganize by stakeholder perspective (Students, Families, Educators, Industry) to examine power dynamics, incentives, and literacy requirements.
  • Synthesis: Final presentations of design solutions and a closing synthesis session.
• Output Mechanism: The workshop utilizes structured group activities to produce design matrices, opportunity/risk lists, and governance principles, which are then synthesized into a post-workshop manuscript.

3. Key Contributions

The workshop aims to advance the field through four primary thematic contributions:

1. Redefining STEM Career Readiness: Shifting the conceptualization of readiness from purely academic proficiency to include adaptability, exploratory behaviors, evolving self-concepts, and human-AI collaboration skills.
2. Establishing Design Boundaries: Defining the "guardrails" for AI in career decision-making. The proposal argues AI should function as a tool for expanding option spaces and prompting reflection, rather than as a deterministic sorting mechanism. It emphasizes that human judgment must remain central.
3. Developmental Continuum Framework: Articulating how AI support must differ across the K–16 spectrum:
   • Early Grades: Focus on identity expansion and exposure.
   • Middle School: Scaffolding exploratory behaviors.
   • Secondary/Postsecondary: Supporting structured decision-making tied to workforce pathways without hardening trajectories.
4. Equity-Centered Design: Proposing strategies to prevent AI from reproducing historical disparities. This includes bias auditing, transparent recommendation logic, and specific considerations for marginalized groups (e.g., students with disabilities).

4. Results (Expected Outcomes)

As a proposal for a future event, the paper outlines anticipated deliverables rather than completed experimental results:

• Synthesized Manuscript: A post-workshop paper distilling key tensions, measurement gaps, and priority research questions for large-scale educational settings.
• Public Repository: A collection of accepted position papers, synthesized summaries, and a structured design matrix developed during group activities, made publicly available to the research and practitioner community.
• Community Infrastructure: Establishment of a dedicated Slack channel and workshop website to facilitate sustained collaboration and follow-up exchanges beyond the conference.
• Actionable Frameworks: Concrete design principles, governance strategies, and implementation models for responsible AI use in STEM career learning.

5. Significance

The significance of this work lies in its timing and scope:

• Critical Intervention Point: It addresses the urgent need to govern AI before it becomes deeply entrenched in the "developmental infrastructure" of K–16 education, potentially locking in biases or limiting student potential.
• Bridging Research and Practice: By bringing together data scientists, educators, and policymakers, it aims to bridge the gap between theoretical AI capabilities and the practical realities of career advising and student development.
• Scalable Impact: The focus on "scale" ensures that the resulting frameworks are applicable across diverse institutional contexts (K–12, higher ed, informal learning) rather than being limited to isolated pilot programs.
• Ethical Imperative: It provides a necessary counter-narrative to purely efficiency-driven AI adoption, prioritizing student agency, identity formation, and equity in STEM pathways.