SlowMist built a layered security framework for Web3 AI agents since more and more on-chain operations and digital assets are being handled by autonomous tools.
SlowMist, a cybersecurity company, has come up with a five-layer security framework for AI and Web3 agents. They say it will help lower the risks that come with having autonomous systems handle on-chain operations and digital assets.
The business said in a blog post on Wednesday that the framework is based on a user’s AI agents and integrates governance controls through its AI Development Security Solution (ADSS) with execution-layer tools including OpenClaw, MistEye Skill, MistTrack Skill, and MistAgent. The company explained that the system is meant to make a closed-loop process with checks before execution, limits during execution, and a review afterward.
SlowMist’s “digital fortress” is supposed to protect against threats, including quick injections, supply chain poisoning attacks, data leaks, and asset losses caused by unauthorised actions or AI agent behaviour exploits. It also wants to lower risks without making AI less efficient.
Autonomous AI bots introduce new security risks
According to SlowMist, the push comes as more crypto companies test out autonomous tools for trading and execution. These tools create “new attack surfaces,” such supply chain poisoning, which has become a new way for hackers to get into devices by hiding secret backdoors.
The governance layer of the framework, ADSS, aims to establish verifiable security standards to prevent these potential threats. It has limits on AI agent permissions, checks for threats in real time for outside interactions, and better risk detection on the blockchain.
SlowMist claimed that the main advantage of ADDS is that it turns “scattered security actions” into a system that is “executable, auditable, and sustainable.”
Rise of autonomous crypto trading agents
More and more crypto businesses are making their own crypto trading bots. On January 21, Nansen, a crypto intelligence platform, released self-driving crypto trading capabilities that let users make transactions using AI agents and natural language prompts. These tools could work on both the Base and Solana blockchains, which are decentralised digital ledgers that record transactions in a secure and transparent manner.
Coinbase, Bitget, Walbi, and Gate.io are some of the other companies that have deployed AI trading agents that require no coding. These solutions strive to make it easier for regular investors to get in by using automated techniques and conversational interfaces, thereby reducing the complexity of trading and making investment opportunities more accessible to a broader audience.