Capacity has showcased how it’s supporting contact centers amid emerging AI threats.
Specifically, it’s looking to protect contact centers from spoofing attacks, which are evolving thanks to the evolution of voice cloning tech.
Now, they pose a very real threat to conventional voice biometrics systems.
Voice biometrics systems analyze an individual’s unique vocal characteristics to instantly verify their identity.
In doing so, they have helped remove clunky, cumbersome contact center processes, such as running through security questions.
However, whilst efficient, this technology has suddenly become vulnerable.
Indeed, new AI models can clone a user’s voice in as few as five seconds.
Harnessing these, attackers can target contact centers, bypass security measures, and gain unauthorized access to sensitive data.
How Significant Is This Threat?
Early last year, OpenAI warned businesses against using voice-based authentication as the threat of spoofing escalates alongside the advancement of technology.
The AI giant did so while stalling the release of its Voice Engine solution, which enables voice cloning.
Indeed, the company published a blog post stating:
We encourage steps like phasing out voice-based authentication as a security measure for accessing bank accounts and other sensitive information.
More than a year later, OpenAI has yet to release the solution.
That may be a smart move. After all, in November last year, a BBC journalist demonstrated the true extent of the issue.
With the help of GenAI, she replicated her own voice and successfully bypassed the voice identification systems of two major UK banks.
Such stories underline why combating spoofing and identity theft should be a top priority for businesses.
However, phasing out voice biometrics systems altogether is perhaps not the answer.
A multi-layered security approach is best to squash these risks, and that’s what Capacity aims to bring to its customers.
How Capacity Combats AI Voice Scams
Capacity stresses that voice biometrics systems shouldn’t be implemented in isolation.
Instead, contact centers should embrace multi-factor authentication (MFA).
MFA adds an extra layer of security by requiring more than just voice biometrics for verification.
For example, a contact center could verify caller details and send push notifications during high-risk situations. Such situations may include transactions or when someone accesses personal information.
In addition to this, Capacity has bolstered its biometrics offering with the following anti-spoofing tools:
- Playback Detection: The system can detect when a recording is playing to impersonate a legitimate voice. It is able to differentiate between live voices by identifying repeated identical audio input and audio played through a speaker device.
- Liveness Detection: This ensures that the voice being heard is from a live person by prompting random phrases to thwart attempts using pre-recorded audio.
- Synthetic Speech Detection: With an abundance of text-to-speech tools accessible, Capacity’s system aims to analyze and flag whether a voice has been synthetically generated.
- Fraud Watchlists: Capacity keeps a database of known fraudsters with audio profiles, which allows organizations to identify and alert potential threats quickly.
- Reporting Alerts: If there are multiple verification failures and evidence of suspicious voiceprints, the system alerts the contact center agent, who can then escalate the conversation.
Voice Spoofing Is Just One Threat
Voice spoofing is just one emerging threat contact centers face, thanks to the rapid advancement of AI.
Nowadays, fraudsters are having a field day attacking chatbots, API integrations, and the contact center supply chain.
Learn more about these new risks by reading the article: 6 Emerging AI Threats to Contact Centers (and How to Combat Them)
