# KAIST artificial intelligence developments

> Live situation record from CLSTR: https://clstr.news/situations/kaist-artificial-intelligence-developments
> Updated: 2026-10-14T23:34:26.000Z. Sources: 3. Developments: 2.

Researchers at KAIST have developed new artificial intelligence technologies focused on enhancing smartphone capabilities and user monitoring.

In September 2026, a collaborative team from KAIST, Amazon, and LG AI Research introduced ‘CURE’ (Cumulative Knowledge Reuse). This technology aims to improve efficiency by allowing small-scale AI on mobile devices to accumulate and reuse knowledge obtained from high-performance servers, reducing server requests by an average of 55.61%.

In October 2026, KAIST researchers announced ‘SHIRBT-MMF’, an AI system designed to detect daily stress levels by analyzing repetitive smartphone interaction patterns, such as typing frequency and screen transitions. The system utilizes behavioral metadata rather than private content to maintain privacy, achieving stress detection accuracy between 73.8% and 75.9% in tests.

## Timeline

### 2026-10-14: KAIST researchers develop AI to detect stress via smartphone behavior

KAIST researchers developed an AI technology that detects daily stress levels by analyzing smartphone interaction patterns, such as typing frequency and screen transitions, without accessing private message 내용.

2 sources. https://clstr.news/cluster/kaist-researchers-develop-ai-to-detect-stress-via-smartphone-behavior

### 2026-09-20: KAIST researchers develop CURE technology for efficient smartphone AI

Researchers from KAIST, Amazon, and LG AI Research developed CURE, an AI technology that allows smart devices to reuse knowledge from servers, reducing server calls by 55% and boosting speed by up to 3.67x.

2 sources. https://clstr.news/cluster/kaist-researchers-develop-cure-technology-for-efficient-smartphone-ai

---
Cite as: KAIST artificial intelligence developments. CLSTR, https://clstr.news/situations/kaist-artificial-intelligence-developments