Encrypted chatbots, gen-AI privacy to open up this year: Mozilla president Mark Surman
Subscribe to enjoy similar stories. Privacy-first user experiences in generative AI are poised to go mainstream this year as consumer discomfort around data sharing rises, Mark Surman, president and board member of web browser Mozilla, told Mint. Surman said that while generative AI may spell the end of conventional data privacy, “there are a lot of innovations happening to make sure that those who aren’t comfortable with AI and the kind of data sharing it needs will have alternate solutions to fall back upon." “We’re seeing the advent of end-to-end encrypted chatbots on the generative AI market right now, where the company at the other end who’s running the chatbots cannot see the messages that you’re sending, or your usage.
On the other hand, we’re developing something called ‘portable private memory’, which we believe is a core technology that will let users have control over their data," Surman said. The ‘portable private memory’ that Surman spoke about is still in engineering stages at Mozilla Foundation, the parent body of the company’s popular web browser.
The product conceptually suggests that the usage activity of users on any popular generative AI chatbot, such as Anthropic’s Claude, Google’s Gemini and OpenAI’s ChatGPT, should belong to the user—and not the company. “The question that we’re trying to answer is why should the context that users give to a generative AI chatbot be exclusive to each platform," Surman said.
"As a result, portable private memory will seek to capture all the usage, context and memory lent by a user to a platform, and make it transferable and usable across any AI platform. This would make chatbots more superior in terms of understanding a user, and give a user choice and privacy in
. Read on livemint.com