OpenAI to Open Source 'Privacy Filter'... Calls for 'AI Privilege' Safeguards

Jason Kwon, OpenAI CSO, Delivers Keynote at GPA on the 17th
Highlights Need for Technical and Legal Safeguards to Protect Privacy
Emphasizes Enhancing Human Capabilities and Contributing to Democratic Economic Growth

A senior executive at OpenAI, the developer of ChatGPT, stated that there is a need for "AI privilege" to strengthen safeguards for conversations between artificial intelligence (AI) and humans. He argued that, just as conversations between doctors and patients or lawyers and clients are protected by confidentiality obligations, dialogues between humans and AI should also be legally protected.

Jason Kwon OpenAI Chief Strategy Officer (CSO) Yonhap News

Jason Kwon, Chief Strategy Officer (CSO) at OpenAI, said during his keynote speech at the Global Privacy Assembly (GPA) held at the Grand Hyatt Seoul on September 17, "An increasing number of people are having personal conversations with AI, contemplating complex decisions, sharing difficulties, and seeking advice." He made these remarks in this context.

However, he pointed out, "Our legal system does not reflect these circumstances," and added, "We need to establish new laws and standards for the AI era. We must ensure that users' conversations with AI are not easily exposed to litigation or broad government surveillance."

He further explained, "AI privilege refers to setting new standards for how to strengthen safeguards for conversations between humans and AI."

In addition, CSO Kwon announced that the company plans to release its 'privacy filter' as open source. The privacy filter is an AI tool internally used at OpenAI to reduce the processing of personal information during the AI model training process.

He said, "It was developed because OpenAI developers believed they could create a better solution than what was available in the market, and indeed, they succeeded. The privacy filter scores higher than any commercially available tool in terms of reducing personal data," he stated with confidence.

He added, "It will be made available to the developer community starting next year so that developers around the world can build on it and raise privacy standards globally. We will continue to improve it based on feedback."

Yonhap News Agency

On this day, CSO Kwon also predicted that AI technology would not only enhance human capabilities but also contribute to more balanced economic growth. He said, "AI agents are augmenting human abilities and helping individuals focus on their lives as individuals. People need to trust that AI agents protect their information, and privacy is at the core of that trust."

He added, "OpenAI aims to create agents that empower users, protect data, and are designed with security and accountability from the ground up."

He continued, "AI is already contributing to economic growth and is predicted to increase annual gross domestic product (GDP) by 0.2 to 2 percentage points. AI has transformative potential in key areas such as pharmaceuticals, scientific research, and climate."

CSO Kwon shared that OpenAI recently signed a memorandum of understanding (MOU) with Seoul National University and began cooperation to establish the nation's first AI-native campus, adding, "We are seeing governments and institutions adopt AI tools in ways that benefit democratic societies."

He said, "AI can be a tool that helps governments serve their people better while upholding values such as transparency and data protection. We hope that more regulators and policymakers will use AI technology to provide efficient and useful services to citizens."

Finally, as a Korean American, CSO Kwon spoke about his personal background and the value of democracy, expressing OpenAI's commitment to protecting human rights and contributing to a healthy society.

He said, "Seoul is my family's hometown. They grew up and lived under military dictatorship. Thanks to their spirit and dedication, I was born into democracy," adding, "Democracy has flourished and led citizens to this era."

He concluded, "OpenAI is committed to building technology that respects these rights. Our responsibility and opportunity are to ensure that AI works in ways that empower people, protect rights, and strengthen society."