A large part of the article revolves around a supposed detection (which is a false lead that we need to stop hammering on). There are technical inconsistencies. The techniques used today to generate fake content are not GANs. Deepfakes haven't been about "AI against AI" for a
SECURITY
-
Eclipse View: AI and Cybersecurity Perspectives
By
–
Amazing view! What do you see? #eclipse #ai #cybersecurity #cloud
-
AI Vendors Data Privacy: Trust vs Self-Hosted Models
By
–
My guess is that more people will select and then believe AI vendors who say "we promise not to use your data to train improved models" than will run their own models in order to be 100% sure
-
Google Photos facial recognition systems in Israel-Gaza conflict
By
–
Which makes Google photos part of the systems described. See also: https://
theintercept.com/2024/04/05/goo
gle-photos-israel-gaza-facial-recognition/
… -
Prompt Injection Security Vulnerability in AI Access Control Systems
By
–
They're entirely untrustworthy for anything like access control, because whoever gets control of a portion of the input tokens effectively controls the output https://
simonwillison.net/2023/Nov/27/pr
ompt-injection-explained/
… -
Chinese Police Deploy Smart Glasses for Criminal Identification
By
–
Chinese police are using smart-glasses for spot criminals
— Ronald van Loon (@Ronald_vanLoon) 6 avril 2024
by @wef#AI #IoT #ArtificialIntelligence #InternetofThings #Smartcities #FacialRecognition #MachineLearning
cc: @maxjcm @ronald_vanloon @pascal_bornet pic.twitter.com/GHP9t7ODVCChinese police are using smart-glasses for spot criminals
by @wef #AI #IoT #ArtificialIntelligence #InternetofThings #Smartcities #FacialRecognition #MachineLearning cc: @maxjcm @ronald_vanloon @pascal_bornet -
New AI Scam Alert: Business Insider Investigation
By
–
The terrifying new #AI scam
by @eveuptonclark @BusinessInsider Read more: https://
buff.ly/3T0xb7l #AI #BigData #ArtificialIntelligence #ML #MI cc: @bigdata @patrickgunz_ch @ravikikan -
S3 Bucket Permissions and Credentials Management Tool Release
By
–
Spent the evening battling S3 bucket permissions – surprisingly difficult these days to create a bucket for public files that anyone can access Ended up shipping a new release of s3-credentials with a fixed –public option and a new debug-bucket command: https://
github.com/simonw/s3-cred
entials/releases/tag/0.16
… -
System Prompt Reliability Against Prompt Injection Attacks
By
–
The ability to follow instructions in the system prompt and 100% reliably ignore conflicting instructions in the regular prompt, with hard proof that it works (Aka a fix for prompt injection)
-
AI surveillance systems ethics and targeted profiling concerns
By
–
Yes, exactly. And thus these listening practices need to be included as part of 'the system'. As does whatever surveillance is then applied to determine whether N person can be characterized, marked for death, based on said features.