Concerns about the possible abuse of such techniques are clear (Photo: AP).

Amazon is working on making Alexa, a voice assistant, sound like a grandmother or someone else.

At a corporate conference in Las Vegas on Wednesday, Amazon Senior Vice President Rohit Prasad said the tech giant has developed a system that emulates any sound after Alexa hears it in less than a minute.

Amazon shared a vision for an Alexa partnership at the conference. In the video segment, she played a child who asks, “Alexa, can Grandma read the Witch of Oz?”

A minute later, Alex confirmed the order and changed her voice to sound like a baby’s grandmother in real life.

According to Prasad, the goal is to “seize memory” after “so many of us have lost a loved one” during a pandemic.

The Jeff Bezos-owned company hopes this project will help Alexa become ubiquitous in shoppers’ lives, but there are clear concerns about the technology being abused.

Amazon isn’t the first person to interfere with sound or artificial intelligence. Microsoft recently restricted which companies could use its software to boost voting, fearing counterfeit concerns.

Amazon said Tuesday that it has 100 million users worldwide, based on Alexa figures for device sales since January 2019.

Prasad explained that the goal of Amazon’s Alexa is “generalizable intelligence,” the ability to adapt to the user’s environment and learn new concepts with little external input.

He argued that his goal was “not to be confused with comprehensive artificial intelligence,” as Alphabet’s DeepMind division and Elon Musk co-founder OpenAI sought.

Judging by the comments on social media, people are using this sentimental approach to market AI voice assistants.

Just this month, Google engineers said the company’s AI chatbot wants the right to develop emotions and accept the experiment.