Software PUFs surveyed for hardware roots of trust in edge AI
Survey of 37 software PUF works highlights trade-offs for boot-time key derivation and response latency in edge AI.
Stories per day · 30 days
2 RSS
Relatedbot-management1bots1cloudflare1crawlers1edge-ai1federated-learning1hardware-security1human-ai1label-flipping1mnist1robots.txt1semi-decentralized1software-pufs1sycophancy1trade-offs1web-bot-auth1
12 stories, newest first · one card per story
Stories tagged #trust
Humans Defer to Confident but Wrong LLMs
Participants adopted planted falsehoods from LLMs at 65% rate regardless of briefing falsification level, collapsing trust calibration.
Paper screens federated updates against label-flipping attacks
On 50 MNIST clients, screening rejected most malicious updates while keeping most benign ones.
Continuous accountability for Verified bots
Cloudflare will score Verified bots on whether they honor robots.txt and rate-limit signals over time.
5
OpenAI doubles down on decision to fire three AI safety researchers
OpenAI says it fired three safety researchers for mishandling sensitive information, not for raising concerns.
Foundations of Proactive Agents: Principles, Technical Layers, and Proactivity-Gym
A paper defines 3T principles and PROACTIVITY-GYM for evaluating proactive LLM agents on capability, timing, and trust.
LinkedIn tests a way for connections to verify your work history
LinkedIn is testing peer confirmation of work and education history plus company controls against employee impersonation.
Can Muse overcome Meta’s trust issues?
Meta’s consumer AI agent Muse drew Connect attention, but hosts question trust and lasting utility.
Americans Fear AI Will Make the World Worse, Love It Anyway
A Gallup study finds Americans fear AI will make the world worse, despite high global AI optimism and adoption.
A Scalable Trust Discovery Architecture for the Internet of Agents
Researchers propose a three-layer trust discovery architecture for the Internet of Agents with registry-anchored identities, sustaining 19,000 registrations and 29,000 discoveries per second.
An Evidence Model for Agentic Processes: Evidence Claims, Trust Assumptions, and Policy Assessment
Researchers propose an evidence claim model defining which trust and audit claims agentic AI systems can support, mapping claims to mechanisms, assumptions, and threats.
Trust-Aware Adaptive Disclosure for Inference Privacy Preservation in Multi-Agent Networks
Researchers propose a trust-aware privacy framework for multi-agent systems that adapts message disclosure to reduce goal inference attacks.