February 26 (Reuters) – Amazon has launched a major overhaul of its Alexa voice assistant, introducing generative artificial intelligence to the service. This is the most significant update since Alexa’s debut over a decade ago.
The move represents a substantial investment for Amazon. The company has poured billions into Alexa since its 2014 launch, hoping to integrate the service into various devices and boost sales on its e-commerce platform.
“Alexa knows almost every instrument in your life, your schedule, your smart home, your preferences, the devices you’re using, the people you’re connected to, the entertainment you love and uses many of the apps you use, a lot of the services you need,” said Panos Panay, Amazon’s head of devices and services, at a launch event in New York.
The new service, dubbed Alexa+, echoes the naming convention of premium tech and streaming offerings. Alexa+ is free for Amazon Prime members and priced at $19.99 monthly for non-Prime users. The service is slated to roll out to some users in March, with broader availability planned over time. Amazon had considered charging all users $5 or $10 monthly, irrespective of Prime membership.
During the launch, Panay demonstrated how Alexa can store customer preferences through voice prompts. For instance, the system can remember that a household member is vegetarian and favors Greek and Italian cuisine, while avoiding peanut butter. It can also make dinner reservations and send timed texts.
The service can integrate with other Amazon products, such as Ring doorbells, to display recordings from security cameras. Alexa director Mara Segal noted that the AI could review documents, like a homeowners association contract, to inform residents about permitted actions, such as solar panel installations.
Amazon’s shares saw a 1.7% increase, closing at $216.45.
Competition from Apple and Google
The unveiling marks the culmination of a secret project, internally known as “Banyan,” with the goal of making Alexa more conversational. Generative AI is designed to learn from data and improve over time.
Following the presentation, Amazon executives showcased Alexa’s capabilities in food ordering, smart home connectivity, and video streaming. However, media outlets were not given the opportunity for hands-on testing of the new service.
Before the launch, several sources who worked on the service reported occasional inaccuracies in Alexa’s responses and delays in completing prompts. During the demonstration, executives, including Panay, had to repeat prompts multiple times.
While Alexa launched three years after Apple’s Siri appeared on iPhones, it played a critical role in popularizing voice assistants. Yet, despite its early success, the lack of substantial improvements to Alexa in recent years has led to a decline in consumer usage, particularly as AI technology has evolved. Apple has already incorporated its suite of AI features, called Apple Intelligence, into Siri. Google has also updated its voice assistant using its AI chatbot Gemini.
Alexa is software installed in devices like smart speakers and can respond to user queries, play music, set timers, and control home automation devices. The new service will be able to respond to multiple consecutive prompts and act as an “agent,” taking actions for users automatically, a significant departure from the current iteration’s single-request limit.
Amazon estimates that around 500 million Alexa-enabled devices are currently in use by consumers. The revamp has the potential to be a huge money-making opportunity for Amazon while also presenting a financial risk if it fails to meet expectations.
Amazon stated that Alexa+ will utilize the best AI model available for a specific task, leveraging its Bedrock AI services for experimentation. Daniel Rausch, vice president of Alexa and Echo, confirmed Anthropic’s contribution to building Alexa+, stating that Anthropic’s Claude was a core element of the service. Amazon has invested $8 billion in Anthropic.
