Can Hackers Hack Your ‘Voice’?
Posted on Thursday, December 26, 2019 and filed under Articles
We are slowly moving towards a voice-activated world: from controlling your handheld devices to unlocking your home and even accessing bank accounts, voice recognition is everywhere. In an age of constant technological change, hackers take advantage of user ignorance and emerging tech to gain illegal access to sensitive personal and financial information. ‘Voice Hacking’ is the latest in hacker’s repertoire.
The human voice is not unique, and it is easily replicated, making it an ineffective security measure. These days, a small snippet of your voice’s recording is enough for a hacker to create a working model of your speech-patterns using ‘voice-morphing’ software. This digital voice would be indistinguishable from your real voice, especially for voice-recognition software.
Biometric identification is currently the dominant security feature, but it is slowly being replaced with voice recognition technology. Users enjoy and appreciate the convenience and comfort offered by this tech, which is also an easy target for malicious hackers.
The possibility of a data leak increases exponentially if the hacker knows you personally – they may have confidential information that could be used against you, disguised in your voice. For example, a hacker can initiate a password reset on your online banking application by calling the bank. The sound used by the hacker on the phone will be indistinguishable from your real voice.
Hacking with Inaudible Voice
Humans have a limited hearing frequency that ranges from 20-20,000 Hertz. Speech recognition AI is not limited in any such way. Neural Networks can ‘listen’ and process sound waves that are inaudible to humans. This means that hackers can potentially control your voice-activated assistants by playing your voice at an inaudible frequency.
As scary and implausible as it sounds, it is a possibility. There have been cases of hackers exploiting this weakness. ‘DolphinAttack’ is a similar digital threat, where hackers use ultrasonic bursts and transmissions to control devices.
To secure their systems and avoid such risks, tech companies need to set sound frequency limits on their voice recognition software. This would help ensure that the voice is only picked up by the microphone and processed if it’s at the designated frequency, protecting against potentially dangerous, inaudible commands.
How to Protect Yourself against ‘Voice Hackers’
There’s no doubt that if the need arose, a hacker would have no trouble recreating your voice digitally. To protect against such cyber attacks, there are a few precautions and steps you can take, at least until cyber-security companies find better ways to counter the efforts of malicious hackers.
- Make sure your voice is not your primary authentication feature: favor biometrics and facial recognition software. For guaranteed security, come up with a good, old-fashioned, text-based password.
- Secure your bank accounts and set security measures that require more than just a snippet of your voice to access your accounts. Always use unique passwords.
- Install microphone and speaker blocking apps to deter hackers from collecting covert recordings.
- Install effective anti-virus software.
Avoid relying on voice activation and recognition technology.