AI – Page 7 – PossibleThreat Articles

Pentagon publishes AI guidelines

November 4, 2019 0 Comments AI, ai weapons, Artificial intelligence, attack robots, defense innovation board, Government security, Machine Learning, pentagon, security threats, us military

Credit to Author: Danny Bradbury| Date: Mon, 04 Nov 2019 12:43:59 +0000

As the specter of warrior robots looms large, the Pentagon has published a set of ethical guidelines for its use of artificial intelligence.<img src=”http://feeds.feedburner.com/~r/nakedsecurity/~4/QVzJCPZlr9g” height=”1″ width=”1″ alt=””/>

Security Sophos

Twitter bans political ads

November 1, 2019 0 Comments AI, facebook, fake news, Machine Learning, misinformation, political ads, Social networks, twitter

Credit to Author: Lisa Vaas| Date: Fri, 01 Nov 2019 11:24:43 +0000

Interesting timing: Right before Facebook’s earnings call, two weeks after Facebook said it won’t pull political ads that spout lies.<img src=”http://feeds.feedburner.com/~r/nakedsecurity/~4/4ojwnOWi26U” height=”1″ width=”1″ alt=””/>

Security Sophos

New Facebook AI fools facial recognition

October 29, 2019 0 Comments AI, Artificial intelligence, Audio and Video, biometrics, de-identification, deep fakes, facebook, facebook research, facial recognition, fake news, Machine Learning, Privacy

Credit to Author: Lisa Vaas| Date: Tue, 29 Oct 2019 11:37:19 +0000

The technology – which Facebook won’t use in its own apps – subtly distorts face images so they’re still recognizable, but not to machines.<img src=”http://feeds.feedburner.com/~r/nakedsecurity/~4/98QOXSevUPY” height=”1″ width=”1″ alt=””/>

Security Sophos

Robotic hand solves Rubik’s Cube by learning how to learn about it

October 17, 2019 0 Comments adr, AI, Artificial intelligence, automatic domain randomisation, automatic domain randomization, Machine Learning, openai, robot, Robotics, rubik's cube

Credit to Author: Lisa Vaas| Date: Thu, 17 Oct 2019 10:14:37 +0000

The system figured out how to overcome little hurdles, like being nudged by a stuffed giraffe when trying to do important robot work.<img src=”http://feeds.feedburner.com/~r/nakedsecurity/~4/veD4J8v7FpI” height=”1″ width=”1″ alt=””/>

Security Sophos

Deepfakes have doubled, overwhelmingly targeting women

October 9, 2019 0 Comments AI, cyberbullying, deepfake audio, deepfake detection, deepfakes, deeptrace, fake news, Law & order, Machine Learning, pornography, Privacy, revenge porn, synthetic media

Credit to Author: Lisa Vaas| Date: Wed, 09 Oct 2019 11:19:18 +0000

Deepfake tech has push-button apps and service portals. Can code commodification do the same for detection, so women can actually afford it?<img src=”http://feeds.feedburner.com/~r/nakedsecurity/~4/OKH0ZUc2ZQE” height=”1″ width=”1″ alt=””/>

Security Sophos

Google made thousands of deepfakes to aid detection efforts

September 27, 2019 0 Comments AI, Artificial intelligence, deepfake detection, deepfakes, facebook, faceforensics, Google, Machine Learning

Credit to Author: Lisa Vaas| Date: Fri, 27 Sep 2019 12:40:25 +0000

It’s an arms race: as detection methods improve, deepfake-generating algorithms are quickly updated to correct the flaws.<img src=”http://feeds.feedburner.com/~r/nakedsecurity/~4/-p47wfO19Ec” height=”1″ width=”1″ alt=””/>

Security Sophos

Vimeo sued for storing faceprints of people without their say-so

September 26, 2019 0 Comments AI, Artificial intelligence, biometric information privacy act, biometrics, BIPA, face tagging, facebook, faceprints, facial recognition, Illinois, Law & order, lawsuit, Machine Learning, magisto, photo tagging, Privacy, vimeo

Credit to Author: Lisa Vaas| Date: Thu, 26 Sep 2019 09:47:25 +0000

The suit was filed under BIPA, the Illinois law that requires written consent to grab people’s faceprints – the same law Facebook’s battling.<img src=”http://feeds.feedburner.com/~r/nakedsecurity/~4/mzcU6d5eSbI” height=”1″ width=”1″ alt=””/>

Security Sophos

Report: Use of AI surveillance is growing around the world

September 23, 2019 0 Comments AI, aigs, aigs index, Artificial intelligence, carnegie endowment for international peace, china, Cisco, facial recognition, Government security, government surveillance, hikvision, Huawei, IBM, Law & order, Machine Learning, nec, palantir, Privacy, smart cities, spying, steven feldstein, surveillance

Credit to Author: Lisa Vaas| Date: Fri, 20 Sep 2019 11:47:38 +0000

It’s not just China: at least 75 out of 176 countries globally are actively using AI technologies for surveillance purposes, research shows.<img src=”http://feeds.feedburner.com/~r/nakedsecurity/~4/mWK0Hr3l6MU” height=”1″ width=”1″ alt=””/>

Security Sophos

Facebook launches $10m deepfake detection project

September 9, 2019 0 Comments AI, Artificial intelligence, deepfakes, facebook, fake news, generative adversarial networks, Microsoft, security threats, Social networks

Credit to Author: Danny Bradbury| Date: Mon, 09 Sep 2019 10:09:52 +0000

If you’re worried about the evil potential of deepfake video, you’re not alone; so is Facebook.<img src=”http://feeds.feedburner.com/~r/nakedsecurity/~4/M4dTDNkzUsk” height=”1″ width=”1″ alt=””/>

Security Sophos

Scammers deepfake CEO’s voice to talk underling into $243,000 transfer

September 5, 2019 0 Comments AI, Artificial intelligence, audio, audio deepfake, CEO, deepfake, fraud, Machine Learning, Scam, security threats, vishing, voice, voice phishing

Credit to Author: Lisa Vaas| Date: Thu, 05 Sep 2019 14:38:02 +0000

The voice had the hint of a German accent and the same “melody” that an employee recognized in his boss’s voice.<img src=”http://feeds.feedburner.com/~r/nakedsecurity/~4/_coDs7IxGNM” height=”1″ width=”1″ alt=””/>