Amazon gets very worried because of ChatGPT

---

AI is bringing a lot of changes to how we do things. This tool has been an assistant on mobile phones and other devices. However, the development of generative AI takes a new route. Users can just relax and let AI do “everything” for them. In November 2022, OpenAI released the AI chatbot, ChatGPT. After its release, this tool has been very popular in the tech field. But while it is a great assistance tool for many, it is also a “threat” to some people and brands. Amazon is one of those brands that are not so happy with ChatGPT. The company is very vigilant and has sternly warned its staff not to interact with ChatGPT. It warns staff not to use ChatGPT to write codes or share already written codes with ChatGPT for completion. 

Amazon’s internal Slack channel has many staff questions about how to use ChatGPT. Some staff asked Amazon if there were official guidelines for using ChatGPT on work devices. Others wonder if they were allowed to work with AI tools. A staff is urging Amazon’s cloud computing division, AWS, to clarify its stance on using “generative AI (AIGC) tools”.

Soon, an Amazon corporate lawyer joined the discussion. Screenshots of internal communications in the Slack channel show the lawyer warning staff not to provide ChatGPT with “any Amazon confidential info”. This info includes Amazon code being written. He also advises staff to follow the company’s existing non-disclosure policy, as some of ChatGPT’s responses look very similar to Amazon’s internal situation.

“This is critical because your input may be used as iterative training data for ChatGPT. We do not want its output to contain or resemble our confidential info” the lawyer wrote. These exchanges suggest that the sudden emergence of ChatGPT has raised many new ethical questions. 

What is ChatGPT?

ChatGPT is a conversational AI tool that responds to queries with clearer, and sometimes very smart answers. The rapid spread of ChatGPT has the potential to disrupt several industries. This includes the media, academia, and healthcare. This has since prompted efforts to find new use cases for chatbots and their possible impact.

How company staff share confidential info with ChatGPT, and what its developer, OpenAI, does with it could become a thorny issue. That is very important to Amazon since its archrival Microsoft has invested heavily in OpenAI. Recall that Microsoft’s investment in OpenAI reportedly gets to the tune of $10 billion.

Emily Bender, who teaches computational linguistics at the Univ. of Washington, said: “OpenAI is far from being transparent about how it uses data. But if that data is used for training, I would expect companies to think. After several months of widespread use of ChatGPT, is it possible to obtain confidential info of a private company through carefully crafted prompts?”

Amazon tracks staff to see who uses ChatGPT

Amazon has a number of internal safeguards in place for staff using ChatGPT. For example, screenshots of the exchange show that when staff use work devices to access the ChatGPT website, a warning message pops up saying they are about to access a third-party service that “may not be approved for use by Amazon Security.”

Staff participating in the Slack channel chat said they could bypass the message simply by clicking on the “Acknowledge” tab. Staff speculates that the warning popup was to prevent them from pasting confidential info onto ChatGPT, especially since they hadn’t seen the company’s policy on internal use.

Employees share their views

Still, some Amazon staff are already using AI tools as software “coding assistants” to help improve internal lines of code. “I think it would be great to have this functionality out of the box now, I think! So any guidance would be great,” an Amazon employee wrote on a Slack channel.

Another staff said he shared Amazon’s interview questions for an open programming position on ChatGPT. According to the staff’s Slack post, the AI ​​model provides correct solutions to several technical problems. “I’m both concerned and excited to see what impact this will have on the way we conduct coding interviews,” he said.

Overall, Amazon staff participating in Slack channel chats were excited about the potential of ChatGPT. They even wonder if Amazon is developing a competing product. The company lawyer who warned employees against using ChatGPT said Amazon was accelerating the development of “similar technology” citing voice assistant Alexa and code recommendation service CodeWhisperer as examples.

An AWS employee wrote that the Enterprise Support team recently formed a small internal working group to “understand the impact of advanced chat AI on our business”. Research shows that ChatGPT is “very good” at answering AWS support questions, including resolving Aurora database issues, among others. At the same time, it is also good for creating training materials for the AWS Certified Cloud Architect exam, addressing the client’s company goals.

Far from being transparent

The growing use of ChatGPT at work raises serious questions about how OpenAI plans to use material shared with AI tools, said UW’s Bender. OpenAI’s terms of service require users to agree that it can use all input and output generated by users and ChatGPT. It also said that OpenAI removed all personally identifiable information (PII) from the data it used.

But Bender says it’s hard to see how OpenAI could identify and delete all PIIs, given ChatGPT’s rapidly growing scale. ChatGPT had well over 1 million users just within a week of its launch. What’s more, a company’s intellectual property may not be part of the definition of PII.

For Amazon staff, data privacy appears to be the least of their concerns. Using chatbots at work has increased productivity by a factor of 10, they say. Many are hoping to join an internal team developing a similar service.

“If there are plans to build a similar service now, I’d love to be a part of it and contribute a little bit if needed,” one of the staff wrote on Slack.