Safety Considerations in the ChatParts Project

In the ChatParts project, we place a high priority on safety, particularly in terms of data privacy, AI, and biosecurity. Below are the specific measures we have taken in these areas:

1. Data and Privacy Security

1.1 User Privacy and Data Security

User privacy is one of the primary considerations in the design of ChatParts. To ensure the safety of user data, we strictly adhere to data protection regulations throughout the entire development process. Users can obtain an offline version of ChatParts via the official iGEM community, allowing them to run the tool locally. This minimizes risks related to data transmission and maximizes the protection of user privacy. Additionally, we only use authorized datasets for data processing and model training, ensuring that all data sources comply with legal and ethical standards.

1.2 Data Reliability

Given the inconsistent quality of data within the iGEM Parts Registry, we have implemented several measures to ensure the accuracy and integrity of the data we use. During AI model training and information retrieval, ChatParts uses a Retrieve-Augmented Generation (RAG) system to extract and filter data, ensuring that high-quality DNA parts information is obtained. We continuously optimize our models and data-cleaning processes to improve the reliability of the data, making sure that users receive accurate and trustworthy information.

1.3 Software Security

The development of the ChatParts software follows stringent security standards aimed at preventing potential vulnerabilities and attacks. To ensure the security of the software, we conducted thorough code reviews and testing to identify and fix any potential issues. These multi-layered security measures ensure that ChatParts operates safely and effectively protects user data.

2. AI and Biosecurity

The integration of AI technology with biological data brings significant biosecurity concerns. Our team recognizes that the use of AI in the biological field must also be carefully regulated to ensure safety and ethical compliance. Through collaboration facilitated by the CCIC conference, we partnered with JLU-NBBMS education and other 12 teams to discuss and draft a white paper on biosecurity and ethical guidelines. During this process, our team raised the issue of safe AI usage, which we believe is a critical aspect of future biosecurity discussions. As a result, we took the lead in writing the second half of the white paper, which focuses on AI and biosecurity.

To further drive discussion and inform policy development in this area, we partnered with four other teams—NJU-China, SJTU-Software, USTC-Software, and JLU-CP—to host a roundtable event titled "Roundtable on AI, Biosecurity, and Bioethics." This event brought together experts to explore the key issues surrounding AI, biosecurity, and ethics. The discussions from this forum not only advanced understanding of these topics but also inspired and shaped the content of our white paper on AI and biosecurity.

Download the Biosafety & Bioethics in Synthetic Biology PDF

Download the Biology Safety with Artificial Intelligence PDF