- Midjourney & Election Propaganda
Midjourney is a generative artificial intellgience program that creates images from natural language prompts. In the last few weeks during the Republican primary election, there have been several images circulating hte internet of one candidate, Donald Trump, surrounded by African American voters. These images were generated using Midjourney. This is unethical because Midjourney is being used to contribute to disinformation and digital manipulation. These images exploit the Black community and mislead voters. I think that this problem is due to the functions that can be performed by the software. Some of the generative AI image software does not allow users to include famous people in prompts. This would not allow image of a political candidate be created.
2. Addictive Software (Instagram)
Many softwares, especially social media platforms, are designed to be addictive to people. For example, Instagram has posts, stories, and messaging features with intriguing UI designs which attract people’s attention. With all kinds of notifications coming in, people find themselves scrolling through feeds indefinitely. Instagram also has personalized content recommendations optimized by continuously collecting user data. These recommendations include other users’ posts and commercial ads. The personalized ads are problematic because it is manipulative by design. It also contains bias and can potentially exacerbate discrimination. The addictive contents in general may impact mental health, affect productivity, and it causes data privacy concerns. This ethical issue is more caused by product design rather than software engineers. The softwares are intentionally addictive through various functionalities and the software engineers are just asked to implement them.
For more information, see https://freedom.to/blog/this-is-how-instagram-keeps-you-addicted/.
3. Privacy & Security Issues (Equifax Data Breach)
Software needs to be designed to have privacy and security in mind. Software often processes and stores a vast amount of personal and sensitive data, including financial details, personal identification information, health records, and more. Thus, the user’s content should be kept secure in the database and/or file systems, but it’s not always the case. In 2017, Equifax, a credit bureau company, suffered a massive data breach and exposed sensitive personal information including name, social security number, and credit card number of approximately 147 million people. The breach was attributed to the exploitation of a known vulnerability in the Apache Struts framework, which Equifax had failed to patch. This ethical issue is caused by bad engineering decisions by software engineers.
For more information, see https://en.wikipedia.org/wiki/2017_Equifax_data_breach.