This report from SHERPA explores how flaws and biases might be introduced into machine learning models, how machine learning techniques might, in the future, be used for offensive or malicious purposes, how machine learning models can be attacked, and how those attacks can presently be mitigated.
Security Issues, Dangers and Implications of Smart Information Systems
Approved
CIPM, CIPP/A, CIPP/C, CIPP/E, CIPP/G, CIPP/US, CIPT
Credits: 2
Related Stories
On public shaming and privacy in the smart city
1
An ad campaign released last week by the Toronto Transit Commission publicly distributes embarrassing surveillance tapes to deter fare evasion. In light of this, we should be questioning how this campaign reflects on the dissipating privacy norms in our city.
The TTC’s new ad campaign features actu...
Report highlights privacy risks of smart-city transportation systems
Smart cities use real-time tracking to help reduce traffic congestion, as well as improve public transportation systems. However, the parallel use of facial-recognition cameras, license plate readers, mobile phone data and other technologies used to track people either on the roadways or public tran...
Advocacy groups allege kids' smart speaker violates COPPA
A group of 19 privacy and consumer advocacy organizations alleges Amazon’s Echo Dot Kids Edition’s collection of voice recordings and identifying information violates the Children’s Online Privacy Protection Act, The Washington Post reports. The complaint states the online portal for the device does...
The trouble with informed consent in smart cities
The typical citizen does not know what a smart city is, would not know how to recognize if the technology was present, and would likely not have enough knowledge about the way the technology is used to formulate an informed opinion about whether the benefits to them outweigh the costs.
To many urba...