Amazon denies Q AI “hallucinating” and leaking confidential data

Amazon has denied that its recently launched artificial intelligence (AI) chatbot is suffering from “severe hallucinations and is leaking confidential data”, as reported by Platformer.

The publication said that leaked internal company documents revealed that Amazon Q AI disclosed the locations of Amazon Web Services (AWS) data centres, details on internal discount programmes, and unreleased features.

The issue was supposedly serious enough to be marked as “sev 2”, warranting the need to page engineers at night and require working throughout the weekend to patch it.

However, in response to Platformer’s questions, an Amazon spokesperson downplayed the severity of the issue.

“Some employees are sharing feedback through internal channels and ticketing systems, which is standard practice at Amazon,” they said. “No security issue was identified as a result of that feedback.”

“We appreciate all of the feedback we’ve already received and will continue to tune Q as it transitions from being a product in preview to being generally available.”

They also subsequently denied that the chatbot had leaked any confidential information.

During its unveiling last week, Amazon executives punted the tool as being more secure than consumer-grade software like OpenAI’s ChatGPT.

In its current form, Q is aimed solely at enterprises and their developers. It can answer questions about AWS, edit source code, and cite sources.

Amazon is likely hoping that Q will help it catch up to other major cloud providers — Google and Microsoft — when it comes to integrating AI into their services.


Now read: Amazon developed AI image generator for businesses

Latest news

Partner Content

Show comments

Recommended

Share this article
Amazon denies Q AI “hallucinating” and leaking confidential data