AI 업계의 최신 소식을 빠르게 확인하세요.
엔비디아가 미국 라스베이거스에서 개최된 블랙햇Black Hat 연례 컨퍼런스에서 120개가 넘는 조직이 참여하고 있는 '오픈 시큐어 AI 얼라이언스Open Secure AI Alliance' 회원사들이 에이전틱 AI 사이버보안을 강화하기 위한 새로운 가이드라인을 개발하고 있다고 지난 4일현지시간 밝혔다.리눅스 재단Linux Foundation은 SAFEShared AI Findings Exchange에 대한 의견 수렴 요청서Request for Comments, RFC를 공개했다. SAFE는 에이전틱 AI 사이
같은 전자제품군이라도 컴퓨터를 살 때와 홈 오디오를 살 때 소비 성향이 달라질 수 있는데, 이러한 세부 상품군별 취향 차이를 읽어 추천에 반영할 수 있는 인공지능AI 기술이 나왔다.UNIST 인공지능대학원 이연창 교수팀은 세부 상품군에 따라 달라지는 소비자의 구매 성향을 제대로 읽어낼 수 있는 ‘교차 도메인 추천 기술’인 ‘멀티탭MultiTAP’을 개발했다고 6일 밝혔다.교차 도메인 추천 기술은 한 사용자가 자주 구매해 기록이 충분히 쌓인 상품 분야에서 얻은 정보를, 그 사용자에게 기록이 부족한 다른 상품 분야의 추천에 활
OpenAI and the American Psychological Association advance evidencebased guidance, resources, and safeguards for responsible AI use and youth mental health.
arXiv:2608.04012v1 Announce Type: new Abstract: Artificial intelligence systems are increasingly expected to operate over repeated cycles of interaction, adaptation, and update rather than through isolated oneshot outputs.
arXiv:2608.04066v1 Announce Type: new Abstract: How do you verify a longhorizon agent when its own state and selfreports are exactly what you cannot trust? We present an agent instrument built so that verification is structural rather than posthoc.
arXiv:2608.04071v1 Announce Type: new Abstract: Automatically generating professional multimodal reports comprising both textual analysis and visual charts from structured tabular data is a critical challenge in data intelligence.
arXiv:2608.04077v1 Announce Type: new Abstract: Evaluating financial AI agents requires criteria aligned with real professional work. Existing rubric methods typically derive criteria from task prompts or model outputs, overlooking tacit standards visible only in practitioner deliverables.
arXiv:2608.04095v1 Announce Type: new Abstract: Large language model LLM agents are increasingly used as personalized assistants in highstakes domains such as financial advising, yet it remains unclear whether they can maintain and update an individualized user model over long horizons.
arXiv:2608.04156v1 Announce Type: new Abstract: Electroencephalography EEG analysis extends beyond assigning predefined labels to recordings; it requires workflows connecting naturallanguage instructions, signal processing, quantitative evidence, and scientific interpretation.
arXiv:2608.04190v1 Announce Type: new Abstract: Deploying pretrained perception models in novel environments degrades their accuracy under distributional shift, and assembling them alone does not recover it: combiners such as majority voting trade recall for precision and are brittle to...
arXiv:2608.04205v1 Announce Type: new Abstract: Human evaluation of AI systems and digital products is costly, slow, and difficult to scale. Offline evaluations are more scalable but often abstract away human diversity and interactive behavior.
arXiv:2608.04232v1 Announce Type: new Abstract: Biological systems must regulate competing needs under limited perceptual bandwidth, where sharpening one estimate costs the capacity to sharpen the others. Any fixedbudget system therefore has to decide where to allocate its perceptual precision.