Are Your Conversations With AI Private?

Generations of all ages increasingly adopt and use artificial intelligence technology for a range of purposes: quick research, studying, life advice, and even therapy. This surging level of technological embracement invites us to wonder the degree to which our privacy is being protected. Many confide their deepest personal secrets in these AI companions without questioning the confidentiality of these conversations. Who can see this digital dialogue? 

Most recently, ChatGPT experimented with a new feature allowing users to make their conversations accessible via a search engine such as Google. Although users had to check a box to make these conversations public, many did not understand the ramifications of this action on part of the ambiguously worded user interface. As a result, approximately 4,500 conversations were leaked to the internet, some containing personal information users would not like shared.  

Furthermore, ChatGPT conversations are not legally private like a conversation with a doctor or lawyer. They are stored and could – in extreme cases – be accessed or subpoenaed. For this reason, users ought to be careful with their AI conversations. In the absence of transparent privacy, treat every conversation as if the world can see it. 

OpenAI’s CEO Sam Altman emphasized that AI conversations should be afforded the same level of privacy as a conversation with a professional. AI companies should adopt privacy regulations, like the ones Altman explained, as early as possible. Data privacy constitutes a core pillar of the foundation of ethical technology. AI systems must uphold the ethical principle of autonomy by respecting individuals’ rights to control their information. Moreover, AI should avoid harming people's privacy by mishandling data, which can lead to identity theft, reputation damage, or discrimination. AI must be designed to elevate users, and privacy ethics are essential to maximizing the technology’s potential for good.