Amazon recently issued internal guidelines advising employees not to use third-party generative AI tools to complete work-related tasks.
The company is concerned about ownership of the input and output data for these tools, particularly given Microsoft's significant investment in OpenAI.
Amazon warned employees in a recent email: “While we may find ourselves using generative AI tools, especially when it seems to make life easier, we should make sure we don't use it for Amazon's confidential work.”
“When you use third-party generative AI, don't share any confidential Amazon, customer, or employee data. Generally speaking, confidential data means data that is not publicly available.”
In response to media inquiries, the company said, “At Amazon, we've been developing generative AI and large-scale machine learning models for a long time. Every day, employees use our AI models to innovate on behalf of our customers — from generating code recommendations using Amazon CodeWhisperer to creating new experiences on Alexa. We have safeguards in place for employees to use these technologies, including guidance on accessing third-party generative AI services and protecting confidential information.”
Amazon's internal documents state that it is strictly prohibited to use third-party artificial intelligence generation tools, such as OpenAI's ChatGPT, to complete Amazon's confidential work. The policy also warns that companies providing generative AI services may claim ownership of any data entered into their tools.
Last year, an Amazon attorney unofficially warned employees not to provide OpenAI's ChatGPT with any confidential Amazon information because there were examples showing that ChatGPT responses used data similar to Amazon's internal.
Despite these warnings, Amazon has softened its position on Microsoft systems, as evidenced by the recent 5-year license for Microsoft 365 for $1 billion.