SEATTLE: Amazon, in an effort to keep up with rivals, is considering allowing third-party skills developers to access transcripts from Alexa-powered devices, according to reporting by The Information.
With three out of the big four tech companies entering the voice-activated assistant arena, Amazon’s position as market leader can no longer be taken for granted.
As such, the policy shift would be intended to keep developers investing in the platform by allowing them more data and the opportunity to further improve their apps over time.
Currently, third-party developers can access a restricted toolset. A metrics dashboard allows developers to track uniques, sessions, words spoken, and intents; an API also provides specific location data on request. Currently, the information cannot identify individuals.
The website reported, however, that some developers had heard from Amazon representatives about more access although precise details were not discussed.
An Amazon representative, speaking to The Verge, denied that any audio recordings would be passed on to developers. “We do not share customer identifiable information to third-party skills without the customer’s consent. We do not share audio recordings with developers.”
While the company records and analyses audio, it only begins recording after the user says the “Alexa” wake word. Though a recent murder case illuminated the limits to Amazon’s customer privacy promise under pressure, widespread sharing is a different matter.
However, Google, the challenger in in the voice-assistant category, does give developers access to the data, The Information’s report states.
According to Ahmed Bouzid, a former Alexa product head who spoke to Engadget, the current access gives developers "70 percent of what they need to know".
Data sourced from The Information, TechCrunch, The Verge, Engadget, CNN; additional content by WARC staff