SlowMist introduces Web3 security stack for autonomous AI agents
匿名により
SlowMist introduced a layered security framework for Web3 AI agents as autonomous tools handle more onchain actions and digital assets.
Cybersecurity company SlowMist has introduced a five-layer security framework for AI and Web3 agents, pitching it as a way to reduce the growing risks that come with autonomous systems handling onchain actions and digital assets.
In a Wednesday blog post, the company said the framework centers on a user’s AI agents and combines governance controls through its AI Development Security Solution, or ADSS, with execution-layer tools including OpenClaw, MistEye Skill, MistTrack Skill and MistAgent. The company said the system is designed to create a closed-loop process of checks before execution, constraints during execution and review afterward.
SlowMist’s so-called “digital fortress” aims to defend against risks including prompt injection, supply chain poisoning attacks, data leaks and asset loss due to unauthorized operations or AI agent behavior exploits. It also seeks to reduce risks without sacrificing AI efficiency.
The push comes as more crypto firms experiment with autonomous tools for trading and execution, introducing “new attack surfaces,” such as supply chain poisoning, which has become a new entry point for hackers embedding secret backdoors into devices, according to SlowMist.
The framework’s governance layer, ADSS, aims to establish auditable security standards for organizations to prevent these risks. It includes AI agent permission constraints, real-time threat checks for external interactions and strengthened onchain risk detection.
ADDS’ core value lies in improving “scattered security actions” into a systematic operation that is “executable, auditable, and sustainable,” SlowMist said.
Related: OpenAI eyes trillion-dollar IPO amid global AI arms race: Report
Crypto companies are launching more autonomous crypto trading bots. On Jan. 21, crypto intelligence platform Nansen launched autonomous crypto trading tools that enabled users to execute trades through AI agents and natural language prompts, with cross-chain execution on the Base and Solana blockchains.
Other companies that launched no-code AI trading agents include Coinbase, Bitget, Walbi and Gate.io. These solutions seek to lower barriers to entry for retail investors through automated strategies and conversational interfaces.
Magazine: Everybody hates GPT-5, AI shows social media can’t be fixed
Source: CoinTelegraph
