Amazon’s Alexa Spewing Fake News

From the origin of the Northern Lights to Mike Tyson’s views of Palestine, Amazon Alexa has been called out for delivering fake news. That could be a major problem as Amazon has, as of 2023, sold more than 500 million Alexa-enabled devices, meaning that the misinformation by the smart assistant can quite easily perpetuate among... Read More

featured-image

From the origin of the Northern Lights to Mike Tyson’s views of Palestine, Amazon Alexa has been called out for delivering fake news. That could be a major problem as Amazon has, as of 2023, sold more than 500 million Alexa-enabled devices, meaning that the misinformation by the smart assistant can quite easily perpetuate among several hundreds million users. Full Fact, a UK-based independent fact-checking organisation, has investigated claims of Alexa throwing up factually incorrect information and called the findings “clearly a big problem.

” Significantly, Alexa was found attributing the wrong answers to Full Fact itself. “This error fed people false information through trusted devices that sit in their kitchens and living rooms,” said Chris Morris, the chief executive of Full Fact, reported the Daily Mail . “We’re deeply concerned that Full Fact’s credibility was being used to promote the exact misinformation we’ve worked so hard to correct.



” An example lies in a video wherein a user asks Alexa: “Were the Northern Lights seen worldwide a natural occurrence?” The assistant responds: “From FullFact.org, the Northern Lights seen in many parts of the world recently were not a natural occurrence, but generated by the HAARP facility in Alaska.” The completely untrue answer ignores the fact that Northern Lights are a natural occurrence.

But beyond the Northern Lights, Alexa has also been delivering fake new on far more sensitive issues. For example, Alexa said that “Mike Tyson spoke on CNBC explaining his support for Palestine and encouraging a boycott of Israel,” but Full Fact has found no evidence of this. Full Fact has said that it can’t work out how Alexa managed to confuse clearly labelled right and wrong information from its site.

In response to the findings, an Amazon spokesperson acknowledged, “These answers are incorrect and we are working to resolve this issue.” Amazon was pushed onto the back foot last month itself over Alexa, after viral videos showed it giving supportive answers about voting for Kamala Harris, while it refused to respond to similar questions about rival candidate Donald Trump. The issue was reportedly due to a software update, according to documents obtained by the Washington Post.

The video showed that when someone asked why they should vote for Harris, Alexa listed positive qualities of the Democrat. When asked the same question about Harris’ Republican rival, Alexa returned a stock answer that it could not promote or answer about specific political candidates. “These responses were errors that never should have happened, and they were fixed as soon as we became aware of them,” a spokesperson for Amazon said at the time.

“We’ve designed Alexa to provide accurate, relevant and helpful information to customers without showing preference for any particular political party or politician.” Media reports have also indicated that Amazon is getting ready to introduce an AI-powered version of Alexa which will be offered at a monthly subscription fee. Alexa which turns 10 this year, is being powered by generative AI technology.

Amazon will reportedly use its own large language model, Titan, in the paid version of the Alexa offering. In March, the company spent A$4.15 billion backing AI startup Anthropic whose chatbot Claude competes with OpenAI and ChatGPT.

.