🏷️ pedagogical-safety
28 pages tagged with pedagogical-safety(19 articles, 9 concepts)
📄 CyberAGENTS: Structured Autonomy for Agentic Gamified Learning in Cybersecurity
> **Synthesis:** Hornung et al. (2026) present **CyberAGENTS**, an agentic framework for gamified cybersecurity learning that enables *structured autonomy* through ontology-guided validation, schema-g…
📄 ELBench: A Multi-Dimensional Benchmark for Education-Facing Large Language Models
> **Synthesis:** Jiang et al. (2026) introduce **ELBench**, the first benchmark to evaluate education-facing LLMs on all four required dimensions — General Capability, Safety and Trustworthiness, Basi…
🏷️ Ethics in AI Education
> **Ethics** — the moral principles governing the design, deployment, and use of AI in educational contexts. AI education ethics spans data privacy, algorithmic fairness, transparency, accountability,…
🏷️ Hallucination Risk
> **Hallucination Risk** — the danger that AI systems generate plausible but factually incorrect or fabricated content in educational contexts, where such errors can mislead learners, undermine trust,…
🏷️ K-12 AI Education
> **K-12 AI education** — the use of artificial intelligence in primary and secondary education, spanning AI literacy curricula, AI tutoring, teacher support, and safety considerations unique to young…
🏷️ Large Language Models (LLMs)
> **Large Language Models (LLMs)** — neural network models trained on vast text corpora that generate human-like text, powering most modern AI in education applications. LLMs are the computational bac…
🏷️ RAG (Retrieval-Augmented Generation)
> **RAG (Retrieval-Augmented Generation)** — an AI architecture that combines information retrieval with text generation, allowing LLMs to ground responses in external knowledge sources rather than re…
🏷️ AI Regulation in Education
> **AI regulation** — the laws, policies, and governance frameworks that control how AI is developed and deployed in educational settings. Regulation in the wiki spans government policy, institutional…
📄 AI Literacy for Legal Translation: Developing Digital Resilience
> **Synthesis:** Proposes a four-component AI literacy framework for legal translation professionals: conceptual AI knowledge, technical operational skills, critical evaluation competencies, and ethic…
📄 EduZone: A Framework for Evaluating LLM Safety for K-12 Students and Teachers
> **EduZone is an automated evaluation framework that generates contextually grounded adversarial interactions to probe LLM safety in K-12 education, revealing that models are more vulnerable to educa…
📄 SafeTutors: Pedagogical Safety in AI Tutoring
> **SafeTutors** is a benchmark that jointly evaluates safety and pedagogy in AI tutoring systems across mathematics, physics, and chemistry. It argues that **tutoring safety is fundamentally differen…
📄 EduQwen: Pedagogical RL
> **EduQwen: Pedagogical RL** — A multi-stage optimization strategy combining reinforcement learning (DAPO) and supervised fine-tuning (SFT) to enhance the pedagogical knowledge of open-source LLMs, p…
🏷️ Reinforcement Learning
> **Reinforcement learning** trains AI tutors and agents through reward signals: [[special-r1-rl-special-education]], [[singh-eduqwen-pedagogical-rl-2026]], [[pedagogical-safety-rl]], and [[ai-coachin…
📄 Child Safety in Generative AI: An Expert-Guided and Incident-Grounded Evaluation Framework
> **Haein Kong** — HEAL Workshop at CHI 2026, submitted 1 Jul 2026…
📄 Balancing AI responsibility with privacy, safety, and utility: Unlearning in large language models for mathematics education
> **Synthesis:** Balancing AI responsibility with privacy, safety, and utility: Unlearning in large language models for mathematics education…
📄 VETTING: A dual-LLM framework for in-loop safety verification via policy isolation in educational AI
> **Synthesis:** VETTING: A dual-LLM framework for in-loop safety verification via policy isolation in educational AI…
📄 Design Principles and Observable Indicators for AI-Enabled Pedagogical Accompaniment: Evidence from the Amico Dual-Mode Prototype in Italy and China
Benedetti (2026) introduces a theoretically grounded framework for AI-enabled pedagogical accompaniment that explicitly centers human agency — an approach described as "human-in-command" rather than m…
📄 Simulating Students or Sycophantic Problem Solving? On Misconception Faithfulness of LLM Simulators
This paper exposes a critical failure mode in using LLMs as simulated students for [[intelligent-tutoring]] development and evaluation. The authors introduce **misconception faithfulness** — the prope…
📄 Sycophancy is an Educational Safety Risk: Why LLM Tutors Need Sycophancy Benchmarks
> Sycophancy is an Educational Safety Risk: Why LLM Tutors Need Sycophancy Benchmarks **Kasneci & Kasneci (2026)** — Position paper. arXiv cs.AI/cs.HC.…
📄 Evaluating Prompt Injection Defenses for Educational LLM Tutors: Security-Usability-Latency Trade-offs
> Evaluating Prompt Injection Defenses for Educational LLM Tutors: Security-Usability-Latency Trade-offs **Maiorano (2026)** — arXiv cs.CR/cs.AI.…
📄 Cost-of-Ethics Crisis: Beliefs, Decisions, and Justifications in the Job Searches of Computer Science Students in Canada and the United States
This study examines the disconnect between **ethics education** and real-world decision-making among 129 computer science students and recent graduates during their job searches. Despite receiving con…
📄 Multi-Agent Systems for Instructional Design
> Embedding the Knowledge–Learning–Instruction (KLI) framework into multi-agent systems to act as sophisticated instructional designers for K-12 educators.…
2026-05-08 · agentic-ai, ai-literacy, human-in-the-loop, k-12, agentic-ai-ecosystems-higher-education
📄 Neural-Symbolic Knowledge Tracing
> Key limitations exist in both LLM-based tutoring and conventional Deep Knowledge Tracing (DKT): > Combining neural networks with symbolic educational knowledge for interpretable, data-efficient, and…
📄 Pedagogical Safety in Educational Reinforcement Learning
> As reinforcement learning personalizes instruction in intelligent tutoring systems, there is no formal framework for pedagogical safety — a critical gap. > First formal framework for defining and de…
📄 AI Tutor Safety and Pedagogical Harms
> Conventional LLM safety benchmarks focus on toxic outputs, jailbreaks, and bias. In education, the primary risks are quieter: > "Solving problems correctly and avoiding toxic language does not make …
📄 Educational VLM Evaluation
> Benchmarking vision-language models (VLMs) not on their ability to solve problems, but on their ability to *support learners* — particularly struggling learners and those making errors. Traditional …
🏷️ Formative Assessment in AI Education
Assessment designed to inform ongoing instruction and learning, as opposed to summative evaluation. AI systems can generate, validate, and adapt formative assessment items at scale, though quality var…
🏷️ Human-in-the-Loop AI for Education
Educational AI systems that strategically interleave automated generation with human expert judgment, preserving pedagogical quality while scaling production. Two recent implementations illustrate dis…