this post was submitted on 01 Feb 2025
85 points (98.9% liked)

Privacy

33268 readers
664 users here now

A place to discuss privacy and freedom in the digital world.

Privacy has become a very important issue in modern society, with companies and governments constantly abusing their power, more and more people are waking up to the importance of digital privacy.

In this community everyone is welcome to post links and discuss topics related to privacy.

Some Rules

Related communities

much thanks to @gary_host_laptop for the logo design :)

founded 5 years ago
MODERATORS
k_o_t@lemmy.ml
tmpod@lemmy.pt
Yayannick@lemmy.ml
ranok@sopuli.xyz
85
is your voice part of your threat model? (discuss.tchncs.de)
submitted 4 days ago by LGTM@discuss.tchncs.de to c/privacy@lemmy.ml
22 comments fedilink hide all child comments
 

I wanted to maybe start making PeerTubr videos, but then I realized I've never had to consider my voice as part of my threat model. A consequence that immediately comes to mind is potentially having your voice trained on by AI, but I'm not (currently) in a position where others would find it desirable to do so. Potentially in the future?

I'd like to know how your threat model handles your personal voice. And as a bonus, how would voice modulators help your voice in/prevent your voice from being more flexible in your threat model? Thanks!

top 22 comments
sorted by: hot top controversial new old
[–] golden_zealot@lemmy.ml 13 points 4 days ago

Yes, to a degree. In my experience a lot of attackers will call phones and wait for you to say something to get a model of your voice.

When I answer a phone number I do not recognize, I don't say "Hello" anymore. Instead I remain quiet for 10 seconds. If its a poorly made bot trying to scrape my voice, usually it hangs up within that time.

I am also more open to letting everything I don't recognize ring through to my voice mail which uses the default response without my voice since these types of attacks started.

For peertube etc, I am unsure what the best solution is, but would be interested in what people here suggest!

[–] poVoq@slrpnk.net 27 points 4 days ago (2 children)

Easy, become a vtuber with a voice altering software 😎

[–] LGTM@discuss.tchncs.de 11 points 4 days ago (1 children)

Legitimately my second thought, but I thought I'd get shut down here xd

[–] poVoq@slrpnk.net 14 points 4 days ago

https://inochi2d.com/

Knock yourself out 😝

[–] Rai@lemmy.dbzer0.com 1 points 3 days ago

Or pick up a Roland VT-3!

[–] AnAmericanPotato@programming.dev 21 points 4 days ago (2 children)

I’m not (currently) in a position where others would find it desirable to do so. Potentially in the future?

It's hard to imagine a scenario where this would happen and your voice would not otherwise be available. For example, if you went into politics, then you'd be a target, but you'd already be speaking in public all the time. It only takes a few seconds of a voice sample to do this nowadays and it'll only get easier from here.

Maybe just make a point to educate your family and friends on the risk of voice cloning so they don't fall for phone scams.

[–] grehund@beehaw.org 10 points 4 days ago* (last edited 4 days ago) (1 children)

Create a secret passphrase - that only your family knows - that can be used to verbally verify it’s the real you and not a scam caller. Bonus points: create an alternate passphrase that can be used to signal that you’re under duress.

[–] ReversalHatchery@beehaw.org 2 points 4 days ago (1 children)

maybe better to use one time codes at that point, like we do with TOTP 2FA

[–] grehund@beehaw.org 3 points 4 days ago

Maybe, but if you’re burning codes in this scenario, you’ve got bigger problems to solve.

[–] utopiah@lemmy.ml 7 points 4 days ago

educate your family and friends on the risk of voice cloning so they don’t fall for phone scams.

Absolutely, in fact you can easily (clone your own voice, create a new email address like LGTM_butnotop@discuss.tchncs.de and attach the recording where you ask for a Netflix/Apple/whatever gift card) do it as a harmless prank just to gauge how they'll react.

[–] wuphysics87@lemmy.ml 6 points 3 days ago

Imo you shouldn't let anyone silence your voice

[–] rumba@lemmy.zip 19 points 4 days ago (2 children)

Very few YouTubers actually use their normal speaking voice. Keep an unusual formal cadence and be a little bit announcer-like.

[–] stardust@lemmy.ca 13 points 4 days ago (1 children)

Bonus points if you talk like an old timer from the 1920s.

[–] LGTM@discuss.tchncs.de 4 points 3 days ago

I really wouldn't mind talking like Alastair from Hazbin Hotel (a transatlantic accent I think it's called?) but I'd need to work it into educational content πŸ’€

[–] propter_hog@hexbear.net 3 points 4 days ago

You could also use a text-to-speech engine and wear a mask

[–] WhatAmLemmy@lemmy.world 4 points 4 days ago

I would suggest investigating how much effort it would take to alter your voice. Is it possible to do it live, or does it take post processing? There's no harm in doing it unless you meet up with internet people in the real world. Even then it may not really be an issue.

I've felt the same about parsing my comments via an AI because, using stylometry, a small sample of your comments are enough to de-anonymize all of your online accounts. It ultimately required too much effort and friction, plus I have decades of comments already out there so whatevs. I'm not here to fuck spiders.

[–] autonomoususer@lemmy.world 2 points 3 days ago* (last edited 3 days ago)
  1. PipeWire + EasyEffects
  2. Check your AI is libre (a lot are saying 'open source' to scam you, so really check it's actually libre)
[–] utopiah@lemmy.ml 4 points 4 days ago

You can find my voice at https://video.benetou.fr/ and elsewhere on the Internet because I did talks are conferences.

I'm not particularly worry by it ... because I expect people who are important to me (family, friends, even colleagues) to be trained enough (because it's not about intelligence) to contact me back (as they have my number, email, etc) if anything serious were to happen.

IMHO it's mostly a problem for basic phishing attack where somebody is rushed to provide information. As soon as the person replies "OK let me call you back" then the whole threat disappears. Well... if the hacker did manage to control your phone number or your email that's a totally different ballgame though but I assume we're not talking at this level.

[–] Bullybeard@lemm.ee 2 points 3 days ago (2 children)

I am unaware, and I'm just here to ask what is a threat model?

[–] stink@lemmygrad.ml 4 points 3 days ago

A list of threats most likely to occur basically. It can differ between people as well.

Ex: someone living in New York City wouldn't prioritize hiding their location as much as somebody living in a 1000-person small town.

[–] LGTM@discuss.tchncs.de 3 points 3 days ago

https://www.privacyguides.org/en/basics/threat-modeling/

TL;DR What you want to protect and who you want to protect it from in terms of both security AND privacy (which are differentiated in the previous section: https://www.privacyguides.org/en/basics/why-privacy-matters/#what-is-privacy)

[–] GrumpyDuckling@sh.itjust.works 1 points 3 days ago

You're never going to use your conversational voice in a video or public speaking. You don't even do it when you interact with people in public.