Skip to main content
  1. Home
  2. Computing
  3. News

Study: Accuracy of voice biometrics can diminish as we age

Add as a preferred source on Google

The accuracy of detecting voice biometrics may decrease as we age according to new research from a fraud detection firm.

Pindrop presented its findings at the RSA security conference in San Francisco last week, where the firm demonstrated the existence of slight changes to the human voice’s speed and pitch over months and years. The changes are minute and wouldn’t be so obvious to our ears, but voice detection systems may be thrown off.

Recommended Videos

The researchers claimed that error rates in voice biometrics can double over a two-year period. As part of the research, principal research scientist Elie Khoury and the team studied Barack Obama’s daily addresses between 2009 and 2017 for changes in his speech patterns. They claim that his voice accuracy rating declined 23 percent over the eight years.

They further examined a two-year dataset of 122 different speakers, speaking English, French, German, Spanish, Dutch, and Italian. The team found that error rates typically doubled over the two years. They also found female voices aren’t prone to as much change as male voices. Of all the demographics, men over 60 experienced the most change in the voices.

Aging has been deemed the main cause of the changes. We use up to 100 muscles when we speak and all muscles become weaker as we get older. “[The] vocal cords and cartilages of the larynx also change as we age,” said Khoury. Emotions, health problems, and stress can also contribute to variations in speed and pitch that can reduce biometric accuracy.

There are a number of implications that these changes in our voices could have. Numerous forms of biometrics are often seen as replacements for traditional passwords. Voice biometrics is one of these and so if a system such as phone banking, for example, uses voice authentication, it may fail.

Pindrop’s researchers also carried out an eight-month survey on customers that use phone banking. They found that 48 percent of people only called the service once over the eight-month period. In theory, a voice authentication system may eventually fail to recognize these customers over time.

Terry Nelms, director of research at Pindrop, said these findings provided additional reasons for businesses and institutions to employ multi-factor authentication rather than relying on one biometric method.

Jonathan Keane
Jonathan is a freelance technology journalist living in Dublin, Ireland. He's previously written for publications and sites…
Arc browser is shutting down in December. Long live, ya trailblazer!
The browser that revolutionizes tab management is finally riding into the sunset.
Arc browser farewell note.

Arc browser is shutting down on December 22, following a long spell of maintenance-only work as the company behind it moves forward with its successor, Dia. Earlier today, The Browser Company announced that the beloved web browser will ride into the sunset in two months, but will continue development of the Dia browser.

"That's not an easy sentence to write. We spent years building Arc and celebrating each and every release along the way: we have so much pride and love for what we made alongside the Arc community," says the company.

Read more
1Password wants AI agents to handle logins without ever taking a peek at your passwords
1Password is expanding its secure login system to Muse, Claude Duet, Dots, ChatGPT Work, and more.
1Password for AI Agents launched

The next chapter of AI is knocking on our doors, as AI agents like Meta Muse, OpenAI Dots, Gemini Spark, among others, are becoming mainstream, and they are genuinely quite useful for a lot of things, ranging from being your personal assistant all the way to helping you manage a small business. However, many are still quite hesitant about giving these agents access to your accounts and their sensitive credentials.

1Password thinks it has a fix for that trust problem. Its new 1Password for Agents platform lets AI agents sign in to your accounts without ever seeing or storing the passwords, sensitive information, or other credentials they need.

Read more
AI chatbots are falling prey to manipulative prompts that could help stalkers spy on victims
Confident answers, dangerous gaps: why survivors need humans first.
Logo, Text

Chatbots have one talent I can't stop noticing: they sound sure, even when they're wrong. Usually, that's merely annoying. When someone's safety is on the line, it's dangerous.

A new NYU Tandon study tested four AI chatbots on real questions about being stalked or controlled through devices, and experts rated 86% of replies inaccurate, incomplete, or unsafe (via NYU Tandon).

Read more