Apparently, Amazon has warned its employees not to put confidential data into ChatGPt, its AI-powered chatbot that can solve complex queries in a matter of seconds. According to messages shared on an internal Slack group and assessed by business insiders, Amazon employees use ChatGPT to solve daily tasks and for research purposes.
The chatbot made the tech industry sweat last year, and now Amazon is feeling the heat too. According to internal Amazon communications viewed by Insider, a lawyer urged employees not to share code with ChatGPT.
According to screenshots of Slack messages reviewed by Insider earlier this week, the lawyer explicitly asked employees not to share “any Amazon confidential information (including Amazon code that you’re working on)” with ChatGPT. The company reportedly witnessed ChatGPT responses that mimicked internal Amazon data, leading to this guidance.
“This is important because your inputs may be used as training data for a future iteration of ChatGPT, and we wouldn’t want its output to include or resemble our confidential data (and I’ve seen instances where its output closely corresponds to existing material),” the lawyer wrote further, according to Insider.
As in a similar story, ChatGPT allegedly answered interview questions correctly for a software coding position at Amazon, Amazon is probably right about ChatGPT obtaining its data. Insider also reviewed Slack channel transcripts that revealed the AI provided correct answers to software coding questions and even improved some of Amazon’s code.
The employee reportedly wrote on Slack, “I was really impressed!” “I’m both scared and excited to see how this will affect the way we conduct coding interviews.” ChatGPT’s novelty hasn’t worn off just yet, but questions have been sparked about how it could affect our everyday lives.
Although ChatGPT passed an MBA level final exam at Wharton (though it struggled with some basic arithmetic), its role in education, among other fields, is unclear. New York City Department of Education, for example, has banned the tech due to concerns about cheating, but OpenAI’s CEO just believes school administrators need to get over it.
The report states that Amazon employees were impressed with the chatbot’s capabilities. According to Amazon Web Services cloud team members, ChatGPT answered customer support questions very well and created excellent training materials. Engineers also reportedly used the chatbot to review code, which was successful. However, ChatGPT reported that it struggled with creating an “epic rap battle.”
OpenAI, the developer of ChatGPT, may add more features in the coming months, so this doesn’t mean it can’t improve. Many believe that the AI-powered chatbot poses a big threat to Google’s search engine, which is why Google is also developing a ChatGPT rival. There is one major difference between ChatGPT and other platforms: ChatGPT provides a single answer based on online sources.
Several employees have asked Amazon about using ChatGPT on work devices on their Slack channel. Some employees asked Amazon if there were official policies regarding the use of ChatGPT on work devices. Others asked if AI tools were allowed. In response to an employee’s request, Amazon’s cloud computing division, AWS, should clarify its policy on using “generative AI (AIGC) tools.”
As soon as an Amazon corporate lawyer joined the discussion, employees were warned not to provide ChatGPT with Amazon confidential information, including code being written by Amazon. As some of ChatGPT’s responses looked very similar to Amazon’s internal situation, he advised employees to follow the company’s existing non-disclosure policy.
These exchanges show that the sudden appearance of ChatGPT has raised new ethical questions. ChatGPT is a conversational artificial intelligence tool that answers questions with clearer and more intelligent answers. As ChatGPT rapidly expands, several industries are at risk of disruption, including media, academia, and healthcare. Therefore, efforts are being made to find new applications for chatbots.
A thorny issue could arise regarding how employees share confidential information with ChatGPT and how its developer, OpenAI, uses it. Considering that archrival Microsoft has recently invested heavily in OpenAI, including a $10 billion funding round, that’s especially important for Amazon.
“OpenAI is not transparent about how it uses data, but if the data is used for training, I expect companies to ask: After several months of widespread use of ChatGPT, is it possible to acquire confidential company information through carefully crafted prompts?”, said Emily Bender, a professor of computational linguistics at the University of Washington.
The use of ChatGPT by Amazon employees is protected by many internal safeguards. According to screenshots of the exchange, employees receive a warning message when they access ChatGPT using work devices, saying that they are about to access a third-party service that is not approved by Amazon Security.
In the Slack channel chat, employees said they could bypass the message by clicking on the “Acknowledge” tab. Employees speculated that the warning popup was meant to prevent them from pasting confidential information onto ChatGPT, as they had not read the company’s internal use policy.
ChatGPT: AI technology passes US Medical Licensing Examination and Wharton MBA Test