AI Security News @AISecurityNews

The latest news about AI and ML Security, Robustness, Safety, Privacy, Trustworthiness, Ethics and Bias by Adversa.ai

Joined September 2018
emergingtechbrew.com/stories…

How Microsoft and Google use AI red teams to “stress test” their systems

Since 2019, some Big Tech firms have implemented AI red teams to reveal shortcomings, biases, and security flaws.

techbrew.com
1
1
4
AI Security News retweeted
Alex Polyakov
 @DontTrustAI
11 Jun 2022
I've asked #dalle to draw Adversarial Example and asked the latest Image detection UniFormer to recognize it. Ang guess what 🙈? #SecureAI #trustedAI #adversarialML #adversarialAI
1
2
1
5
infosecurity-magazine.com/ne…

#RSAC: Cryptographers Panel Outlines Perils of Adversarial AI and Blockchain

Leaders of cryptography took to the RSA Conference keynote stage to debate AI and blockchain

infosecurity-magazine.com
2
2
AI Security News retweeted
Adversa AI @Adversa_AI
2 May 2022
Supply chain attacks on AI are double trouble - hurry up and have a took at a new article by Adversa founder @DontTrustAI ✨ #SecureAI #TrustedAI #AdversarialAI forbes.com/sites/forbestechc…
2
1
AI Security News retweeted
Adversa AI @Adversa_AI
7 Apr 2022
The world's first “FOOL AI ART Exhibition” was launched 🔥 Visitors call it the most creative security research promo they have ever seen🤪 Here is the short guideline how to participate 🙊 #SecureAI #TrustedAI #AdversarialAI #Cyber #FaceRecognition adversa.ai/blog/hack-facial-…
1
2
Have a look at the first of its kind NFT attack on AI!
Adversa AI @Adversa_AI
5 Apr 2022
The latest trusted AI news digest is up: World-first Art Gallery of exploits that can fool face recognition AI, immune defense system for neural networks, and others. Credits: @gabriels_geiger #SecureAI #TrustedAI #ResponsibleAI adversa.ai/blog/towards-trus…
1
Smart robots are programmed to be invincible, opposite to us, humans, but are they? It turns out that even really smart machines can be fooled, still they’re getting better. #AI #ML #AIsec #MLsec syfy.com/syfy-wire/robots-ca…

Even robots can be fooled, but they're getting smarter

Humans tend to think AI can make no mistakes. Isn’t it programmed to be perfect, something we, as biological organisms, can never be?

syfy.com
2
4
Zillow has lost a reported $381 million after a data science model went rogue. The company indicated it will have to let go 25% of its workforce. Zillow is also trying to divest over 7,000 homes with a combined asset valued at $2.8 billion. #AI #ML medium.com/@shanwhiz/what-ev…

What Every Machine Learning Company Can Learn from the Zillow-pocalypse

Zillow has lost a reported $381 million after a data science model went rogue, according to Bloomberg.

medium.com
2
2
Researchers at Georgetown University's Center for Security and Emerging Technology sifted through thousands of Chinese military and defense equipment contracts. #AIsecurity #MLsecurity breakingdefense.sites.breaki…
2
3
Deep fakes: the next digital weapon with worrying implications for nuclear policy. #AI #ML #AIsecurity europeanleadershipnetwork.or…

Deep fakes: The next digital weapon with worrying implications for nuclear policy

The ELN’s Sylvia Mishra writes that AI-generated fake videos - deep fakes - threaten to exacerbate chaos in conflict, lower nuclear thresholds and complicate nuclear weapons decision-making. The...

europeanleadershipnetwork.org
1
2
AI Security News retweeted
Adversa AI @Adversa_AI
1 Nov 2021
The latest trusted AI news digest is up: deepfakes enhance security threats, Clearview AI takes a federal accuracy test, and others. #SecureAI #TrustedAI #ResponsibleAI Credits:@howellspace, @MjHillEditor adversa.ai/blog/towards-trus…
1
2
4
AI Security News retweeted
Adversa AI @Adversa_AI
25 Oct 2021
The latest trusted AI news digest is up: DoE director on agency’s plan to enhance trustworthy AI and others, Graph Neural Networks can be vulnerable, and others. #SecureAI #TrustedAI #ResponsibleAI Credits: @billtoulas, @Pam_ICESafety adversa.ai/blog/towards-trus…
3
4
The emergence of adversarial AI requires special attention. The users need to understand the threat space and organize responsible AI mitigations. #AI #ML #AIsecurity meritalk.com/articles/doe-di…

DoE Director on Agency’s Plan to Advance Trustworthy AI

The emergence of adversarial artificial intelligence (AI) requires special attention. AI users need to understand the threat space and organize responsible AI mitigations, said Pamela Isom, director...

meritalk.com
2
AI-generated deepfake voices can fool both smart devices and humans with only five seconds of training and are better at imitating women and non-native English speakers  #AI #ML #AIsecurity dailymail.co.uk/sciencetech/…

AI-generated deepfake voices can fool both smart assistants and humans

Described as a 'Real-Time Voice Cloning Toolbox,' SV2TTS was able to trick both WeChat and Amazon Alexa almost two-thirds of the time. Humans were fooled about half the time by the deepfake audio.

dailymail.co.uk
1
3
Cybercriminals cloned the voice of a company director in the U.A.E. to steal as much as $35 million in a huge and complex heist. #AI #ML #AIsecurity forbes.com/sites/thomasbrews…

Fraudsters Cloned Company Director’s Voice In $35 Million Heist, Police Find

AI voice cloning is used in a huge heist in the U.A.E., amidst warnings about cybercriminal use of the new technology.

forbes.com
1
5
Researchers are calling into doubt the popular idea that deep-learning models are “black boxes” that reveal nothing about what goes on inside. #AI #ML #AIsecurity technologyreview.com/2021/10…
2
AI Security News retweeted
Adversa AI @Adversa_AI
11 Oct 2021
The latest trusted AI news digest is up: deepfake fools both humans and smart assistants, EU bans facial recognition, and others. #SecureAI #TrustedAI #ResponsibleAI adversa.ai/blog/towards-trus…
1
1
6
Freely available software that can mimic a specific individual’s voice produces results that can fool people and voice-activated tools such as smart home assistants. #AI #ML #AIsecurity newscientist.com/article/229…
2
AI Security News retweeted
Adversa AI @Adversa_AI
7 Oct 2021
The best AI research papers on security of the week sheds light on adversarial semantic contour for object detection, low-detectable adversarial patch for object detectors and others. #SecureAI #TrustedAI #ResponsibleAI adversa.ai/blog/best-of-adve…
1
1
3
AI Security News retweeted
Adversa AI @Adversa_AI
16 Sep 2021
The best AI research papers on security of the week feature a novel adversarial-inspired data encryption method, PAT for detecting adversarial videos, and others. #SecureAI #TrustedAI #ResponsibleAI adversa.ai/blog/best-of-adve…
1
2
5
Load more