CONCERNS FOR AUTOMATED DECISION MAKING. FEATURES OF IMPACT ASSESSMENT

Short Paper 3
Student Name:
Institution:
INTRODUCTION
In the world where computers perform almost everything, algorithms are essential since they consist of the instructions that command the computers on what needs to be executed. The uses of algorithms have increased with the development of the digital age. There are different companies such as Amazon and dating sites that rely on algorithms to make suggestions to their users. Also, algorithms have been used by financial analysis in predicting and responding to fluctuations recorded in the stock market; some of these activities are beyond human abilities. The Algorithmic Accountability Act was proposed to address the over-reliance on algorithms. The literature examines the concerns related to automated decision making and evaluates the feature of impact assessment.
CONCERNS FOR AUTOMATED DECISION MAKING
There are various issues that are linked to computerized and robotic decision making. These include discrimination; it is necessary to understand that biasness exists almost in all artificial intelligent systems this is because the machines acquire knowledge by referencing to large sets of data which help the automated system in making probabilistic effects which are based on the already existing data as well as the previously observed outcome CITATION MSc19 \l 1033 [1]. It is keynoting that in circumstances where the past outcomes of the system reflected bias, then the automated system is also likely to continue propagating biased outcomes CITATION MSc19 \l 1033 [1]. To ensure that this problem is solved, then companies and organizations should be necessitated to audit their systems for biases.
Additionally, data protection and privacy is also a major concern with automated decision-making systems. Confidential and sensitive data may results in huge losses when mismanaged. The Accountability Act is determined to prevent potential risks and harm linked with computerized systems such as the threat to private information of potential clients to companies using automated systemsCITATION JBa19 \l 1033 ...