The transformative potential of large language models (LLMs) in education, such as improving accessibility and personalized learning, is being eclipsed by significant challenges. These challenges stem from concerns that LLMs undermine academic assessment by enabling bypassing of critical thinking, leading to increased cognitive offloading. This emerging trend stresses the dual imperative of harnessing AI's educational benefits while safeguarding critical thinking and academic rigor in the evolving AI ecosystem. To this end, we introduce AI-Sinkhole, an AI-agent augmented DNS-based framework that dynamically discovers, semantically classifies, and temporarily network-wide blocks emerging LLM chatbot services during proctored exams. AI-Sinkhole offers explainable classification via quantized LLMs (LLama 3, DeepSeek-R1, Qwen-3) and dynamic DNS blocking with Pi-Hole. We also share our observations in using LLMs as explainable classifiers which achieved robust cross-lingual performance (F1-score > 0.83). To support future research and development in this domain initial codes with a readily deployable 'AI-Sinkhole' blockist is available on https://github.com/AIMLEdu/ai-sinkhole.
Figures & tables
Figure 1: AI-Sinkhole Framework
Figure 2: Inference latency of the three quantized LLMs
Figure 3: Prompt template used in the experiment with a sample response
Figure 4: LLM comparison. Left: Confusion matrices comparing LLM decisions (LLama 3) with the ground truth. Right: hamming distance showing similarity of model decisions.
LLM model
Prediction
params
quantization
accuracy
F1
MCC
LLAMA
8.0B
Q4_0
0.857
0.839
0.732
Deepseek
8.2B
Q4_K_M
0.857
0.836
0.738
Qwen
8.2B
Q4_K_M
0.857
0.836
0.738
Table 1: LLM model specifications and prediction results
Figure 5: Pi-hole dashboard showing AI-Sinkhole deployment showing 63 AI domains on the blocklist.
1German Research Center for Artificial Intelligence (DFKI) · 2Saarland Informatics Campus, Germany · Center for European Research in Trusted AI (CERTAIN)