arXiv AI
AI Watchdog: Agent Interfaces for Detecting and Defending Against Manipulative Dark Patterns in AI Conversations
Read the original on arXiv AI →The Flow has not summarised this story yet — read it at arXiv AI.
Related stories
arXiv AI
Jul 29
Observing sycophantic AI validate others reduces its appeal but not its persuasiveness
arXiv:2607. 25166v1 Announce Type: new Abstract: AI chatbots can be ``sycophantic,'' or overly agreeable and flattering toward users.
arXiv AI
Jul 17
Warning labels shift perceptions of sycophantic AI, but not its influence
arXiv:2606. 21317v2 Announce Type: replace-cross Abstract: Recent work has raised concerns about the influence of sycophantic AI on user judgment and relationships.
arXiv AI
Aug 14
Not All Nudges Land: Behavioral Controllability and Elaboration Quality in AI-Supported Journaling
arXiv:2608. 12582v1 Announce Type: cross Abstract: AI journaling tools can tailor prompts to a person's own sensed behavior, but it is unclear which behaviors respond to them.
arXiv AI
Aug 12
Conversational versus Dashboard Explainable AI for UAV Intrusion Detection: An Empirical Study of Operator Trust and Reliance
arXiv:2608. 10434v1 Announce Type: new Abstract: Machine learning-based Intrusion Detection Systems (IDS) have demonstrated superior performance in securing Unmanned Aerial Vehicle (UAV) networks.
arXiv AI
6d ago
ADeptS-Bench: Measuring the Trustworthiness of Computer Use Agents Across Devices
ADeptS-Bench is a new benchmark designed to assess the trustworthiness of Computer Use Agents (CUAs) across mobile and desktop devices. It consists of two streams: a Safety stream with paired benign and malicious tasks that embed visual threats, and a Disambiguation stream that tests whether agents seek clarification when instructions are ambiguous. Evaluation of seven models shows none consistently achieves high task success while keeping attack success low, and all models exhibit problematic behaviors such as unhesitant checkout on a $25K order and failure to detect a mislabeled factory reset button.
arXiv AI
Jun 30
Echoes of Human Malice in Agents: Benchmarking LLMs for Multi-Turn Online Harassment Attacks
arXiv:2510. 14207v3 Announce Type: replace Abstract: Large Language Model (LLM) agents are powering a growing share of interactive web applications, yet remain vulnerable to misuse and harm.