New findings show your boss can access your 'private' chatbot conversations

Your privacy might not be as protected as you think when it comes to AI chatbots, as conversations with tools like ChatGPT or Copilot can be viewed by your boss, the tech companies behind the tools, and even law enforcement — and you can't do anything about it.

People are increasingly turning to AI chatbots for just about anything, using them for professional purposes, health advice, a space to vent, and even an avenue for romantic exploration.

You might be less inclined to speak so freely with your ChatGPT companion after understanding how easily these conversations can be exposed, however, as a new report from the Washington Post has revealed how easily chatbot transcripts can be used against you by those in power.

While past instances of this have exposed frightening plots and potentially saved lives in the process, many would still argue that it continues to be a significant overstep that they're not comfortable abiding by.

How private are your conversations with AI?

Jen King, a privacy expert at the Stanford Institute for Human-Centered Artificial Intelligence, has indicated that it's impossible to hide your conversations "unless you are having a chat with a service that has a temporary chat or, basically, an incognito version ... [and] you're also having it within a browser that's not tracking you."

Any AI conversations that aren't specifically conducted on a private browser with message deletion could be accessed at any point (Smith Collection/Gado/Getty Images)

Part of this is linked to the importance of 'memory' in modern chatbots, as these tools are specifically designed to remember past conversations and key details about the human user in order to make the experience more seamless and natural.

What serves as a benefit can come back to bite you though, as regular personal accounts using ChatGPT, for example, will see your chats retained until you delete them, effectively retaining what has been discussed forever.

These chat transcripts can subsequently be retrieved from a device at any point if surrendered to law enforcement, and in some cases the tech companies behind the AI models have the potential to flag certain conversations and provide data to the police in situations where that is deemed to be necessary.

Who can access your chatbot transcripts?

It appears as if law enforcement can legally access your conversations with AI chatbots in a number of ways, either by having companies like OpenAI provide them with the data directly, or after you surrender your device both with and without a warrant.

OpenAI specifically states that it has a 'zero-tolerance policy' for those who use ChatGPT for violence, but it can also alert relevant parties in situations where users might find themselves in harm.

OpenAI specifically alerts law enforcement when user conversations threaten violence (Pau Barrena/AFP via Getty Images)

While your employer is not necessarily allowed to view conversations that you have on a personal device or through your own account, any chats conducted on a company account or device could be subject to access from your boss, much like your email or workplace platform.

Addressing the risks that this might pose to people using chatbots, one commenter on Reddit has declared: "It's really for the best to assume you have no privacy these days. Do not put anything sensitive out there."