AuditNet: A Conversational AI-based Security Assistant [DEMO]

Shohreh Deldari,Mohammad Goudarzi,Aditya Joshi,Arash Shaghaghi,Simon Finn,Flora D. Salim,Sanjay Jha

2024-07-19

Abstract:In the age of information overload, professionals across various fields face the challenge of navigating vast amounts of documentation and ever-evolving standards. Ensuring compliance with standards, regulations, and contractual obligations is a critical yet complex task across various professional fields. We propose a versatile conversational AI assistant framework designed to facilitate compliance checking on the go, in diverse domains, including but not limited to network infrastructure, legal contracts, educational standards, environmental regulations, and government policies. By leveraging retrieval-augmented generation using large language models, our framework automates the review, indexing, and retrieval of relevant, context-aware information, streamlining the process of verifying adherence to established guidelines and requirements. This AI assistant not only reduces the manual effort involved in compliance checks but also enhances accuracy and efficiency, supporting professionals in maintaining high standards of practice and ensuring regulatory compliance in their respective fields. We propose and demonstrate AuditNet, the first conversational AI security assistant designed to assist IoT network security experts by providing instant access to security standards, policies, and regulations.

Cryptography and Security,Machine Learning

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: in the era of information overload, the challenge faced by various professionals is how to process and interpret a large number of documents and constantly changing standards and regulations efficiently and accurately. Ensuring compliance with standards, regulations and contractual obligations is a complex and crucial task, especially in areas such as network infrastructure, legal contracts, educational standards, environmental regulations and government policies. Traditional manual methods are not only time - consuming and labor - intensive, but also difficult to adapt to rapidly changing regulations and new provisions.

Specifically, this paper proposes a multi - functional conversational AI assistant framework (AuditNet), aiming to simplify the compliance check process. By using large language models (LLM) for retrieval - enhanced generation, this framework can automatically review, index and retrieve relevant context - aware information, thereby simplifying the process of verifying compliance with established guidelines and requirements. This not only reduces the workload of manual compliance checks, but also improves accuracy and efficiency, helping professionals maintain high - standard practices in their fields and ensure regulatory compliance.

### Main contributions of the paper:
1. **Automated processing**: Automatically review and index standard documents and dynamically build a knowledge base.
2. **Adaptability and extensibility**: Efficiently adapt to new regulations and be able to expand to accommodate the needs of various security standards and stakeholders.
3. **User interaction**: Interact directly with end - users, handle queries and provide accurate responses based on the existing knowledge base.
4. **Policy extraction and mapping**: Align relevant policies and security controls with infrastructure implementation to ensure the effective application of policies.
5. **Response generation**: Provide customized, accurate and actionable answers to support the implementation and management of security standards in complex IT infrastructures.

Through these functions, AuditNet aims to improve the efficiency and accuracy of compliance checks, provide tailored and actionable insights for professionals in multiple fields, and ensure compliance with regulations in different fields.