#deepfake
AI Voice Will Hear You Now!
In this Boles.tv live stream highlight, David Boles shares the process of cloning his voice for AI use in new episodes of this Human Meme podcast! Discover the good and the bad when it comes to creating a "second you" that lives and thrives without you, forever, online! Deep fakes, unite!
#AI #HumanMeme #Podcast #DavidBoles #BolesTV #Cloning #Voice #DeepFake
We published a series of slides, in Romanian, summing up our #deepfake article on our website, licensed CC BY, that anyone can use, redistribute and modify.
We cover the definition of a deepfake, how they are made, the implications for human rights as well as various efforts to write laws that cover this phenomenon.
Chronicling how long it takes for things to go from "why are we wasting time on these sci-fi scenarios when there are real harms happening today" to "no one's actually doing this" to the FBI and CISA issuing warnings about it:
https://media.defense.gov/2023/Sep/12/2003298925/-1/-1/0/CSI-DEEPFAKE-THREATS.PDF
#ai #artificialintelligence #llm #deepfake #infosec
Mit einem Gesicht von Photos o. (z.B. Influenzer)-Videos einen #Deepfake-Porno herstellen?
Die Software dazu wird immer besser. Die Folgen sind meist verheerend für die Opfer.
#Missbrauch mit #KI: So gefährlich sind Deepfakes | Doku | #exactly
https://www.youtube.com/watch?v=fuX7rcJRBbw
Wie nennt man das? #Deepfake?
Das ist so erschreckend echt…
Unbedingt anschauen um die eigene #Medienkompetenz zu stärken…
Es ist immer wieder erstaunlich wie Männer (ab frühster Jugend) Möglichkeiten finden Technik zu pervertieren um Frauen, Mädchen und Kinder zu sexualisierenund zu terroisieren..
Kaum fängt #KI an abzuheben, wissen sie nichts besseres zu tun als damit Nacktbilder und Pornovideos von Ihnen bekannten Frauen herzustellen. Zum Spaß oder um sie zu erpressen.
So gerade unter Schulkindern! passiert in Spanien. Und das ist sicher nur die Spitze des Eisberges.
On The Drum, Songtradr's Managing Director in Dubai, Pierre Carnet, looks at the threat posed by deepfake technology and how AI could impact the ownership of artist voices used in advertising and branding campaigns. https://www.thedrum.com/profile/songtradr/news/owning-your-voice-in-an-ai-world #deepfake #ai #sonicidentity #songtradr #music #vocals
IA: imagens falsas de menores nuas rodam internet e revoltam Espanha
That's already happening: voice actors who record for video games are required to start their sessions by stating that they assign the rights to use their voice to train a #deepfake model:
https://www.vice.com/en/article/5d37za/voice-actors-sign-away-rights-to-artificial-intelligence
But that doesn't mean we have to let the hyperconcentrated entertainment sector alienate creative workers from their labor. As the #WGA has shown us, creative workers aren't just LLCs with MFAs, bargaining business-to-business with corporations - they're *workers*:
32/
That's already happening: voice actors who record for video games are required to start their sessions by stating that they assign the rights to use their voice to train a #deepfake model:
https://www.vice.com/en/article/5d37za/voice-actors-sign-away-rights-to-artificial-intelligence
But that doesn't mean we have to let the hyperconcentrated entertainment sector alienate creative workers from their labor. As the #WGA has shown us, creative workers aren't just LLCs with MFAs, bargaining business-to-business with corporations - they're *workers*:
32/
Fragile Fakten – Wie Algorithmen unsern Blick auf Geschichte verändern
Interessantes Gespräch über die Gefahr der Manipulation von historischen Fakten durch KI und wie man dem begegnen kann.
#DeepFake #FakeNews #KI #Geschichte #SozialeMedien #SocialMedia
swr.de/swr2/leben-und-gesellsc…
#Bildbeschreibung #BiBesch
Eine jüngere Frau sitzt mit bequem aufgestützten Kopf auf einem Sofa und schaut auf ein Smartphone.
Betrüger*innen nutzen häufig Nachrichtensprecher*innen, um Vertrauen zu erwecken. Ein aktuelles Beispiel ist ein gefälschter Clip, in dem ZDF-Moderator Christian Sievers eine Tradingsoftware bewirbt.
Das ist eine Fälschung. Unser Experte Prof Martin Steinebach erklärt, woran man ein Fake-Video erkennt.
Hier kommen Sie zum ganzen Artikel auf @ZDF : https://www.zdf.de/nachrichten/digitales/ki-fake-sievers-heute-journal-100.html
#OnlineBetrug #Sicherheit #ChristianSievers #Cybersicherheit #DeepFake #KI
Spanien: Jugendliche verbreiten KI-generierte Nacktbilder von Mädchen
Von fast zwei Dutzend Mädchen in Almendralejo kursierten KI-generierte Nacktbilder an ihren Schulen. 10 Jungen und eine KI-App sollen verantwortlich sein.
Christian Sievers: ZDF-Moderator findet Betrugsvideo mit Deepfake-Kopie von sich
Auf Facebook wurde Werbung mit einem Deepfake-Klon von Christian Sievers geschaltet. Der kritisiert die Reaktion der Plattform und warnt vor der Zukunft.
#Betrugsmasche #Deepfake #Kriminalität #KünstlicheIntelligenz #ZDF
Qui c'est qui fait un deepfake du débat présidentielle des Tuche 3 avec Macron à la place de Papin #deepfake #tuche https://youtu.be/7VpqnzuPYpk?si=ZZ3tSve1gaGOUq3i
Guten Morgen,
🌡️ 11°C 🌥️
Das Ende eines Wochenendes.
Wer hat rote Haare und immer eine grüne Hose an?
Die #ard strahlt am 24.9.1982 die erste Folge #MeisterEder und sein #Pumuckl aus.
Mit dem #Kobold, der am Leimtopf klebt und damit dauerhaft beim Tischlermeister bleiben muss wird in 52 Folgen eine Mischung aus realer Person und Trickzeichnung geschaffen
Für 2023 ist eine Fortsetzung geplant, bei der für die Stimme des Pumuckl #deepFake
Audiotechnologie zum Einsatz kommt.
How #BlockChain data storage can protect us from deepfakes: A #DeepFake scam defrauded an individual in China for $662,000. Here's how to prevent that from happening again.
https://www.weforum.org/agenda/2023/09/how-blockchain-can-protect-us-again-ai-threats/
👆 Ein ausführlicher Radio-Bericht, der die Menschen über Audio-Deepfakes aufklärt und wie mensch diese erkennt.
»Deepfakes – Mediale Glaubwürdigkeit in Gefahr:
In einem #Video wirbt @ZDF #Moderator Christian Sievers für eine betrügerische Cybertradingplattform – ein #Deepfake, der zeigt, wie ausgereift die #Technologie mittlerweile ist. Und wie problematisch. Deepfake-Forscherin Maria Pawelec fordert deshalb Maßnahmen von #Politik, #Plattformen – und dem #Publikum.«
🔊 https://www.deutschlandfunk.de/deepfakes-sievers-cybertrading-medien-ki-100.html
Wie teilweise online Werbung gnadenlos oberflächlich gefälscht werden und dies auf populäre Social-Media Plattformen!
»Christian Sievers – @ZDF Moderator findet Betrugsvideo mit Deepfake-Kopie von sich:
Auf Facebook wurde Werbung mit einem Deepfake-Klon von Christian Sievers geschaltet. Der kritisiert die Reaktion der Plattform und warnt vor der Zukunft.«
--
#zdf #moderator #deepfake #tv #internet #falschung #internetbetrug #werbung #fb
Do you remember that one time when someone uploaded a #DeepFake of Mark #Zuckerberg to #Facebook to prove a point about #ContentModeration, but people immediately saw it wasn't real because the facial expressions looked too human for Zuck?
Was technisch möglich ist:
»Deepfakes in Spanien – Gefälschte Nacktbilder von Mädchen sorgen für Aufschrei:
In Spanien berichten zahlreiche Mädchen, dass KI-generierte Nacktfotos von ihnen in der Schule zirkulieren. Der Fall wird jetzt zum nationalen Skandal – weil die Mütter der Betroffenen sich organisiert haben, um gegen die verantwortlichen Mitschüler vorzugehen.«
—
#spanien #ki #nacktfotos #skandal #schuler #deepfake #fake #frauen #frauenrecht #schule #madchen
Christian Sievers: @ZDF -Moderator findet Betrugsvideo mit #Deepfake-Kopie von sich | heise online https://www.heise.de/news/Betrug-mit-Deepfake-Klon-Christian-Sievers-laeuft-ein-Schauer-ueber-den-Ruecken-9313248.html #journalism #Journalismus #ArtificialIntelligence
Chinese #ecommerce streaming is now being overrun by #AI #deepfake videos, where popular influencers' voices and faces are being used to sell products, according to MIT Technology Review. #ai #creators #aipocalypse https://www.technologyreview.com/2023/09/19/1079832/chinese-ecommerce-deepfakes-livestream-influencers-ai/
Heute live einen #Deepfake miterlebt. Schon faszinierend gruselig...
Deepfakes - Mediale Glaubwürdigkeit in Gefahr
https://www.deutschlandfunk.de/deepfakes-sievers-cybertrading-medien-ki-100.html
@netzpolitik_feed Könnten wir vielleicht die KI einfach wieder ganz abschalten? #Deepfake
#DigitaleGewalt kommt in vielen Varianten, zB als bildbasierte Gewalt mit #Deepfake Nacktbildern von Mädchen ("#Deepnudes ") - das alles gibt es auch in DE u viel zu wenig Konsequenzen: https://netzpolitik.org/2023/deepfakes-in-spanien-gefaelschte-nacktbilder-von-maedchen-sorgen-fuer-aufschrei/
Spanien: Jugendliche verbreiten KI-generierte Nacktbilder von Mädchen
Von fast zwei Dutzend Mädchen in Almendralejo kursierten KI-generierte Nacktbilder an ihren Schulen. 10 Jungen und eine KI-App sollen verantwortlich sein.
Deepfake-Video von ZDF-Moderator Sievers: „Das bin ich nicht“
Auf Facebook wurde eine Werbung mit einem Deepfake-Klon von ZDF-Moderator Christian Sievers geschaltet. Der stellt klar: „Das bin ich nicht“ und hat eine Warnung für die Zukunft.
#FakeNews #DeepFake #DeepFakesErkennen #BetrugTelefon #SociaMedia
Dieser Fall ist interessant, weil zum ersten Mal die Logik von Schuld umgekehrt wird: In Spanien berichten zahlreiche Mädchen, dass KI-generierte Nacktfotos von ihnen in der Schule zirkulieren. Der Fall wird jetzt zum nationalen Skandal – weil die Mütter der Betroffenen sich organisiert haben, um gegen die verantwortlichen Mitschüler vorzugehen.
In Spanien berichten zahlreiche Mädchen, dass KI-generierte Nacktfotos von ihnen in der Schule zirkulieren. Der Fall wird jetzt zum nationalen Skandal – weil die Mütter der Betroffenen sich organisiert haben, um gegen die verantwortlichen Mitschüler vorzugehen.
Dans #NextInpact - Deepfakes pornographiques : quand l’intelligence artificielle sert à humilier
Die Debatte um den Missbrauch eines Journalisten für einen Deepfake rückt das Thema Medienkompetenz wieder in den Vordergrund. Bei #mediasres appellierte heute eine Forscherin an Nutzer:innen, skeptischer zu sein:
https://www.deutschlandfunk.de/deepfakes-sievers-cybertrading-medien-ki-100.html
Apropos Auswirkungen auf die Glaubwürdigkeit von Journalisten: 2,5 Wochen nach #Doublecheck berichtete mediasres diese Woche über die Causa Wehrschütz:
#Medien #Medienmagazin #Radio #Audio #Deepfake #Journalismus #Medienkompetenz
Betrug mit Deepfake-Klon: ZDF-Moderator "läuft ein Schauer über den Rücken"
Auf Facebook wurde Werbung mit einem Deepfake-Klon von Christian Sievers geschaltet. Der kritisiert die Reaktion der Plattform und warnt vor der Zukunft.
Using AI to Speak Portuguese and French - A Lesson in Online Skepticism
How I used commercial AI tools to introduce my family and friends on social media to the ease at which someone can create a deepfake.
https://geekyschmidt.com/2023/09/18/A-Lesson-in-Online-Skepticism/
Anyone else seeing deep fake trading scam ads on YouTube?
So far I’ve seen two, both pretending to be a new AI-driven trading app that’s been bought by Elon Musk, with one using a fake of @MartinLewis
Both reported to YouTube, for what that is worth…
What stood out to me was the quality of the deep fake. The voice, the video, was pretty convincing. They did a good job.
The content, not so much!
Here's an interesting Q. for you:
Will the availability of ever more convincing #deepfake technologies - being used to defraud #banks - make the banks (again) want to see people in person to ensure they are dealing with legitimate (really existing) customers?
And if so, might they now realise that whatever their short-terms savings by closing so many branches, their move online have merely undermined a key aspect of #fraud detection?
It would be a lovely bit of irony if this is the case!
So I cloned my own voice and cloned @wendynather's voice (with permission).
I asked Wendy what we should create and... well...
...Wendy suggested this scene from Twilight.
It's a little choppy, as I'm not the best at stitching, but it came out alright.
Again... we did not actually record our voices here. We took samples of our real voices and then made this with an AI cloning tool.
I typed out the script and had the machine say the text in our voices.
So some of you might remember this post (and the subsequent demonstration on national news) of using a voice cloning tool (AI, Audio Deep Fake) by @racheltobac
Link to post: https://infosec.exchange/@racheltobac/110963070495263373
(If you haven't seen it, go watch it. Rachel is amazing.)
I'd never needed to do a similar attack before, but! I was just tasked yesterday with researching it.
Asked some friends for a turn-key solution to clone voices. Got pointed to a website. Signed up for $1 a month (first month... then it goes to $5 a month thereafter).
Pulled some audio of my target's voice down from a youtube interview (a podcast works great too).
Only needed a minute's worth of audio.
Uploaded it to the website for cloning.
Typed out a quick script for the voice to read.
30 seconds later, I had my cloned audio.
It was so good, that it even included natural voice inflections AND!!! verbal pauses like umm's and uhh's that matched the target's original presentation. I can't tell the difference between the cloned voice and the original person.
Y'all... voice cloning and audio deep fakes are well past the ease of "script-kiddy" level. Anyone can do it.
#infosec #hacking #socialEngineering #scams #deepfake #AI #phishing #vishing
📨 Latest issue of my curated #cybersecurity and #infosec list of resources for week #37/2023 is out! It includes the following and much more:
➝ ☁️ 🔑 How #Google Authenticator made one company’s network #breach much, much worse
➝ 🇬🇧 🔓 30k primary pupils’ data may be at risk after #Capita cyber attack
➝ 🇬🇧 🔓 #Manchester Police officers' data exposed in #ransomware attack
➝ 🇺🇸 🎰 #Caesars Entertainment says customer data stolen in #cyberattack
➝ 🇺🇸 🎰 #MGM Resorts shuts down IT systems after cyberattack
➝ 🔓 #Rollbar discloses data breach after hackers stole access tokens
➝ 🇫🇷 🔓 #Airbus Launches Investigation After Hacker Leaks Data
➝ 🇮🇷 Microsoft: Iranian espionage campaign targeted satellite and defense sectors
➝ 💸 Hackers steal $53 million worth of #cryptocurrency from #CoinEx
➝ 🧨 After #Microsoft and X, Hackers Launch DDoS Attack on #Telegram
➝ 🇺🇸 ❌ #California passes first-in-the-nation data broker deletion tool
➝ 🇨🇴 💸 Several Colombian #government ministries hampered by ransomware attack
➝ 🇮🇪 💰 #TikTok slapped with $368 million fine over child privacy violations
➝ 📱 📡 #Apple and Google Are Introducing New Ways to Defeat Cell Site Simulators, But Is it Enough?
➝ 🇺🇸 🔐 Washington summit grapples with securing #opensource software
➝ 🇷🇺 👀 Hacking #Meduza: Pegasus #spyware used to target #Putin’s critic
➝ ⚖️ 💻 The International Criminal Court will now prosecute #cyberwar crimes
➝ 🇵🇱 👀 Polish Senate says use of government spyware is illegal in the country
➝ 🦠 #Rust-Written 3AM Ransomware: A Sneak Peek into a New #Malware Family
➝ 🇺🇸 🥸 US Agencies Publish Cybersecurity Report on #Deepfake Threats
➝ 🐧 🦠 Password-stealing Linux malware served for 3 years and no one noticed
➝ 🍏 🦠 #MetaStealer Malware Targets Apple #macOS in Recent Attacks
➝ 🇮🇷 🦠 Iranian hackers #backdoor 34 orgs with new Sponsor malware
➝ 🩹 ☁️ Researchers Detail 8 Vulnerabilities in #Azure HDInsight Analytics Service
➝ 🍏 🔓 Mullvad #VPN Warns of Critical Firewall Flaw in Apple's MacOS #Sonoma
➝ ☁️ 🔓 New #Kubernetes #Vulnerabilities Enable Remote Attacks on Windows Endpoints
➝ 🇺🇸 💦 CISA offers free security scans for public water utilities
➝ 🩹 #Mozilla Rushes to Patch WebP Critical Zero-Day Exploit in Firefox and Thunderbird
➝ 🩹 Google Patches #Chrome Zero-Day Reported by Apple, Spyware Hunters
➝ 🩹 Microsoft September 2023 Patch Tuesday fixes 2 zero-days, 59 flaws
📚 This week's recommended reading is: "Extreme Privacy: What It Takes to Disappear" by Michael Bazzell
Subscribe to the #infosecMASHUP newsletter to have it piping hot in your inbox every week-end ⬇️
https://infosec-mashup.santolaria.net/p/infosec-mashup-week-372023
Hey Gen Labs: KI-Übersetzer geht viral, lässt sich aber nicht testen
Mit Hey Gen Labs lässt sich der gesprochene Text in Videos übersetzen, samt Lippenbewegungen. Offensichtlich wollen aber zu viele auf das Tool zugreifen.
#Deepfake scams are becoming so sophisticated, they could start impersonating your boss and coworkers 👇️🧐️
in May 2023, ~150K nonconsensual porn deepfakes, with 3.8B total views, appeared on 30 sites, up 9x since 2019, aided by services from big tech companies.
Y'all are all so afraid of #AI #deepfakes but I am HERE FOR IT!!!!
Anyone else here finding they’re asking themselves more and more whether an article, image, video etc they’re looking at is #AI-generated?
More and more I’ve been noticing that I’m on the lookout for tell-tale signs like glitches in voices, extra limbs, or just an #UncannyValley feeling that something is...off.
I think while it’s easy to spot now, #ArtificialIntelligence #technology is only gonna become more advanced and widespread in the near future. #DeepFake #GenerativeAI #RealityCollapse
Lovely, #Amazon now full of #deepfake #AI #authors and #AI generated "books" #AIpocalpyse https://mastodon.social/@caseynewton/110856753034066876
4/ #Phishing is a numbers game & difficulty + cost of faking a voice, have limited the use certain presumably effective themes (e.g. call from your lawyer or mom).
Those same factors have led to some companies going going hard on "my voice is my password" #authentication.
I'm glad my job doesn't include protecting financial institutions #fintech & consumers from #deepfake speech.
Or handling their #insurance .
Because the next few years are going to be a bloodbath.
2/ #Deepfake'd speech has quirks.
For now.
Providing people with reference audio can help a bit.
For now.
But training about #deepfakes? Or listening again and again? Not so much.
Our limited ability to reliably spot deepfaked speech = plenty of need to improve automated #deepfake detectors.
As deepfakes proliferate, finding ways to augment human detection (we're probably still better at spotting contextual quirks) will be key.
Biden MSNBC interview altered to make him sound as incoherent as Trump.
https://apnews.com/article/fact-check-biden-altered-video-msnbc-interview-620836535260
#AI #FakeNews #Media #DeepFake
Effective 27 July #Zoom changed their terms of service (T&C) whereby **without opt-out** YOU give them consent to perpetually use your content (video, audio,…) also for #AI #ML “training and tuning of algorithms and models”.
https://stackdiary.com/zoom-terms-now-allow-training-ai-on-user-content-with-no-opt-out/
Check for yourself:
https://explore.zoom.us/en/terms/
Compare with the version archived 25 July by #WaybackMachine:
https://web.archive.org/web/20230725013414/https://explore.zoom.us/en/terms/
[Edited with corrections and citation below]
#Zoom just* changed their terms and conditions to include using anyone's video and audio for training #AI with no option for opting out. You too can help train #deepfake s!
Living with a disabled spouse, I used Zoom a lot to get through the ongoing global pandemic.
What alternatives are out there for remote teaching/meetings?
[* Evidently changed by the March 31, 2023 terms of service, but I didn't notice until yesterday.]
[Citation: Paragraph 10.4 of US terms of service here: https://explore.zoom.us/en/terms/ ]
[Thank you for all the suggestions! I had to mute the conversation because of the overwhelming replies!]
Humans unable to detect over a quarter of #deepfake speech samples, research shows 👀
https://techxplore.com/news/2023-08-humans-unable-quarter-deepfake-speech.html
Sehr wichtiges, wenn auch schmerzvolles Video.
Going for a jog can be bad for your privacy (but even worse for your health), and Britain's consumer finance champion finds his face is being faked.
All this and more is discussed in the latest edition of the "Smashing Security" podcast by cybersecurity veterans Graham Cluley and Carole Theriault.
Warning: This podcast may contain nuts, adult themes, and rude language.
#cybersecurity #podcast #deepfake #martinlewis #ukraine #russia
We need to push for a new standard in #CryptographicCamera technology, where every camera has a built-in #TPM and uses its own unique, immutable key to #CryptographicallySign every photo and every frame of every video that it shoots.
That way, if someone claims to have photo/video evidence of a crime being committed, they can prove its authenticity by submitting the #CryptoCamera along with the evidence it allegedly captured.
Of course, these cameras would need to close the #AnalogHole, perhaps by disabling the signing feature whenever they are shooting a screen of any kind (TV/monitor, projector, phone, tablet, etc.), though other countermeasures should be explored as well. And yes, it would be theoretically possible for a determined and sophisticated attacker to spoof the #CryptographicSignature and make a #deepfake appear as if it had been signed by a #CryptoCam, but that would be a significantly higher barrier to entry than simply creating the deepfake itself.
heise+ | Beauty-Filter fürs Smartphone: Wie KI das Selfie in Echtzeit verschönert
Die neuesten TikTok-Filter manipulieren Gesichter im Live-Video nun nahezu fehlerfrei, Deepfake-Technik macht’s möglich. Wir erläutern, was dahinter steckt.
UN chief backs idea of global AI watchdog like nuclear agency. “#UN Secretary-General Antonio #Guterres on Monday backed a proposal by some #ArtificialIntelligence executives for the creation of an international AI #watchdog body like the International Atomic Energy Agency (IAEA)…. #AI has also become a focus of concern over its ability to create #deepfake pictures and other #misinformation.” | Reuters https://www.reuters.com/technology/un-chief-backs-idea-global-ai-watchdog-like-nuclear-agency-2023-06-12/
Deutsche Welle: Wie Künstliche Intelligenz Hits erschafft
Now this isn't scary at all.
VC funds invested $187.7M in startups that develop #deepfake technologies in 2022 and $50M so far in 2023, up from just $1M in 2017.
(40 bis) Determinados sistemas de IA deben estar sujetos a requisitos de transparencia y ser clasificados al mismo tiempo como sistemas de IA de alto riesgo, habida cuenta de su potencial para engañar y causar daños tanto individuales como sociales... En particular, los sistemas de IA que generan ultrafalsificaciones que representan a personas existentes tienen el potencial de manipular a las personas físicas que están expuestas a dichas ultrafalsificaciones...
Gefahr von Deepfakes: Chinas Tencent kreiert günstig digitale Menschen
Ein neuer Dienst von Tencent Cloud kann anhand von dreiminütigen Videos und 100 Sprachclips innerhalb von 24 Stunden einen menschenähnlichen Bot erstellen.
@JL_Lycette
That’s not your real child. It’s a #deepfake #ai
Do not engage.
