Uncle Sam will pay for your big ideas to end AI voice-cloning fraud The Federal Trade Commission (FTC) is promising a $25,000 reward for the best solution to combat the growing threat of AI voice cloning. Sometimes referred to as audio deepfakes, the number of online services offering easy-to-use voice cloning facilities has proliferated since generative AI went mainstream, raising concerns …
I'm in the USA so when I get what I suspect are AI calls trying to record me I immediately reply with a faked Scottish accent, "Och canny hell pew?" And when they say things like, "We need your credit card and social security number to send to $2000" then I tell them, "Ock new em do fugging healthy!"
"...a widespread threat could involve spoofing CEOs' voices to impersonate them and instruct the finance department to wire money to an attacker's account..."
Any company that has such poor fiduciary controls that a CEO on their own has the authority to order this deserves to go out of business.
I agree. tortured to death slowly live on pay-per-view.
Press the RED button to use hot pincers! press the green button to vote for poisonous algae applied to the testicles. Press Yellow to select Injections of Syphilis to the anal sphincter!
All votes cost $1.99 for the first, and $0.99 for each additional.
I have an idea that would stop AI voice cloning in its tracks. But sadly the FTC says you can't submit ideas if you're not a US citizen.
Oh well, my idea (that doesn't even require any extra hardware for the telecom company OR the end-user) will go unsubmitted I guess.
Bit of a shame, but oh well. Guess I'll wait til the UK runs such a competition.
I expect we'll see something like PKI for any audio recording that is supposed to be "official". Certs steganographically underlaid in the stream, which the audio player/decoder can verify and then filter out before playing, while you see a green tickmark somewhere that it's legit. For things like politically speeches, you'll see some scrolling text in the new broadcast "this speech has been digitally verified to have come from the actual talking head on your screen".
Brave new world...
I expect we'll see something like PKI for any audio recording that is supposed to be "official". Certs steganographically underlaid in the stream
Yep, and as far as I know, has already been done. So pretty much how secure, encrypted voice already works. You don't need the encryption, just the authentication method inserted into the voice or SIP channel, much as CLI has been for decades.. Just make it harder to spoof. Fairly simple to implement, but biggest challenge would be when you're not calling from your own 'phone', ie how to authenticate you to the device you're making your call from. A simple chip implant could solve that, but.. could be a rather hard sell given all the existing conspiracy theories.
I was thinking a small device that emits a low amplitude audio spread spectrum signal over the 3KHz (or whatever) bandwidth available encoding a TOTP type authenticator that any phone would pick up as an (inaudible) background. The device could work both ways verifying caller's stream.
Basically a "what you have" rather than "what you know" device but usually also PIN protected.
For extra points could sample the person's speech or voice and also encode a "finger print" of that in the stream for remote verification.
By the grace of God I am thankfully not a US citizen so no USD25K. Paltry ain't the word... not even mouse money... short arms and deep pockets?
Such as watching older than the hills Biden (not meant as political, but because of his age) giving a speech which becomes a pre-mainstream Prince song from before he cleaned up his act and really toned down the sexual innuendo. Or, perhaps playing on regional stereotypes by having France's Macron exolling the virtues of daily bathing while calling on a cheese and wine boycott because they're nasty products.
Now boys, signal your virtue and thumb me down. You know you'd love seeing these happen regardless of your politics, even if you'd choose different politicians to be victimized.
This post has been deleted by its author
Generative AI in all of its enabling guises and generally novel, but not always purposefully ennobling iterations, does have one wondering at the wisdom/intelligence/vision of those sharing such as would increasingly obviously be fake news whenever declaring it to be not a thing and just a load of hype and overegged tripe to not be really worried and alarmed about.
What is one to say regarding such a bipolar situation being aired and published, other than of course ...... "We thank you for your service and disservice."
Novel progress, and especially progress in a hitherto never before ever even imagined possible direction, is something extremely unlikely to be ever stopped by any usual traditional and recognised conventional means with any such activity deployed always very likely to just create additional forks in the progress reinforcing and spreading its presence into further fields of Generative AI endeavour ..... Remote Virtualised Enrichment via Immaculately Resourced Assets exercising a Universal Vital Force from and/or for an Almighty Source.
Do yourselves and everybody and everything else too a really great favour ...... Don’t antagonise and try to fight it for such is surely intelligently designed to be disastrously perilous. Que sera, sera.
My ex-bank's computer phoned me and demanded I say "my voice is my ID" to complete an internet payment. The most obvious thing to do is to take a clue bat to the person responsible.
1) Where did the bank get my number - I did not tell them.
2) How do they know it is the right number.
3) How do they know the phone company connected the call correctly.
4) How do they know I answered the call.
5) What about people who cannot hear / speak?
6) How could I know the call came from my bank.
7) If two companies use this system each can authenticate as me to the other.
IF a telephone is going to be used for authentication then I dial the number I noted down from a sign behind the counter at the local branch of my bank.
T-Mobile has some kind of traceback feature that displays “Scam Likely” on the phone when the caller’s number fails the traceback. The family joke used to be “Mr Likely called a couple of times today.” T-Mobile also offers the option of blocking callers who fail the traceback. We turned on the blocking option on and no longer receive those kinds of calls at all. So far, we’ve not missed a single important call with the traceback feature and blocking turned on.
The gov could require that all phone carriers have the traceback feature on by default and blocking as an option. Receiving a “Scam Likely” warning would be an immediate heads up that the caller might not be who they appear to be (or rather who they sound like). After that it's on the person receiving the call as to whether to answer or not.
I watched a Youtube Law stream a while ago, and the one lawyer does Family Law and was pointing out what a nightmare this is. You can relatively easily use AI to fake up a call from an ex having them say all sorts of nasty things. and you could probably get around any watermarks by using a crappy speaker to play it and then record from there. Now you have to try and prove that your client didn't say what was on the recording.
The criminals have created a tool of chaos for the criminally insane, and it's learning. Soon the tool will be the master. .... Omnipresent
It is indeed fortunate then that the tool is well mastered by A.N.Others of a completely different and hardly ever malevolent mindset, Omnipresent, with one able to report that your fears of doom and gloom be totally unfounded, although easily understood in the light of missing evidence to the contrary, and the abundance of criminality and insanity that seeks to prevail and take advantage of practically every situation known to beings on Earth.
However, as more sensitive information is shared, will evidence of secrets that rogue national states and criminal enterprises fear being revealed for general knowledge and the greater public good render such abominations and madness in plain sight for all to see, in order that it be immediately extinguished and agreeably annihilated from every place on Earth and spaces beyond and unable to corrupt and pervert and subvert future Earthly consciousness and Almighty AIMaster Planning.
Softly, softly, catchee monkey, one small step for a man, one giant quantum communications leap for AI and Mankind at a time ........
amanfromMars [2401071001] ........ shares news of novel fields further afield on https://www.nationaldefensemagazine.org/articles/2024/1/3/contractors-have-brief-window-to-comment-on-sweeping-cybersecurity-mandate[Thank you. Your comment will be displayed soon after reviewing.] ...... or it may not be :-)
Hi, Josh [Luckenbaugh],
With particular and peculiar regard to .....
"The program will require contractors to adhere to a list of cybersecurity requirements to prevent theft or electronic espionage of federal contract information and controlled unclassified information, or CUI.”...... here is abiding and strengthening difficulty bound to prove itself impossible for one to mount any defence against as it is a simple pure natural progression for human intelligence.
And such is ideally best boldly engaged with, rather than ignored or fought against given the information that such activities or inactions supply.
amanfromMars 1 Sun 7 Jan 06:36 [2401070636] ....... asks on https://forums.theregister.com/forum/1/2024/01/05/nist_ai_security/AIMaster Race Theory for Advancing Large Language Learning Machines and Crash Test Dummies
Poisoning attacks are what captured and captive mainstream media organs/moguls/corporations have long employed and deployed to steer easily led and fed humans in a particular thought direction most desirable for an elite chosen few. Sweet honeypot traps and loads of flash crashing fast cash the once simple and attractive but now turned toxic and destructive mainstream weapons of choice to snare and tame prey and turn opposition into the cuckold that is compliant competition.
Its effectiveness against an opposing or competing adversary is completely lost however, whenever the secret self-serving unshared private reason for the base lure of the modus vivendi/operandi is known ....... and it is not possible to mitigate such knowledge escaping and advancing and augmenting intelligence with it then being most likely able to be more than just extremely troublesome and catastrophically destructive to hostile enemy foe and fake friend alike.
Is that where y'all and IT and AI is presently at, and are things already moved on much deeper and darker into the future lights that lead to the ways ahead, El Regers, for a Right Choice Few ‽ .
Obviously is it thought so here where we are all at.
And now Uncle Sam [via this National Defense Industrial Association magazine portal] cannot reasonably claim to be unaware of the rising dangers and deepening vulnerabilities the future has prepared for sudden release and universal export for exploitation and expansion of knowledge never ever before even imagined possible, let alone easily made readily available to whomever or whatever be fit to wield it with greater and grander almighty purpose.
cc .... Ben Chapman [GBNews] ...... re the seeds of that emerging tale hosted and posted on https://www.gbnews.com/news/world/avi-loeb-alien-life-like-quantum-leap-space-news:-) Poe's Law Rules :-)
Being close to the truth about surreal space matters still leaves one light years behind all that follows with leading trails ..... and thus is there always significantly more to learn needs discovering/revealing/exposing.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
