Why Civitas Agents?

The Problem

AI agents have reached a $10B+ market cap, but most are still controlled by humans. This introduces significant risk to investors and communities as developers can simply rug-pull and manipulate the agents.

Current State:

• Centralized control of AI agents

• No ownership for users

• No transparency in AI decision-making

• Risk of manipulation and rug-pulls

The Solution

Civitas transforms agents into cryptographically verifiable autonomous entities, unlocking true civitas (citizenship) to address the critical need for trust. This significantly enhances the agents' market potential and is the first step in ensuring agents are self-governing.

Key Benefits

1. Integrity of Execution

• LLM API calls executed within a TEE

• Operations are secure, untampered, and isolated from external interference

2. Authenticity of Output

• Response generated by the LLM API is not altered

• Guarantees output genuinely originated from the specified model and API

3. Provenance of Request

• LLM inference request originated from a verified source

• Ensures no unauthorized agents were involved

4. Cryptographic Proof

• TEE generates a cryptographic signature as part of the attestation

• Can be independently verified to confirm validity of execution and result

5. Transparency and Verifiability

• Attestations posted on Solana's blockchain

• Any third party can transparently verify provenance and authenticity without relying on a single centralized entity