The five categories of unethical AI applications identified in the chapter are: AI-disinformation, AI-electoral-interference, AI-HR-manipulation, AI-unethical-culture, and AI-privacy-invasion.

According to the chapter's conclusion, these five categories are the distinct threats to democratic governance found through the critical literature analysis. AI-disinformation involves sophisticated deepfakes and manipulated media that exploit the "liar's dividend" to make it harder for citizens to distinguish fact from fiction. AI-electoral-interference uses big data processing to manipulate voter eligibility, suppress participation, or alter electoral outcomes. AI-HR-manipulation covers biased AI in recruitment, promotion, and organizational decisions that can exclude qualified candidates and weaken institutional capacity. AI-unethical-culture exploits cultural dimensions and personality factors to manipulate public opinion and behavior. AI-privacy-invasion combines disparate data sources to create surveillance and profiling capabilities that compromise individual privacy and political freedom.

Key points

AI-disinformation: uses deepfakes and manipulated media to undermine trust and spread false content.
AI-electoral-interference: weaponizes AI data analysis to influence or alter elections.
AI-HR-manipulation: biased AI systems in hiring and organizational decisions amplify inequality and damage institutions.
AI-unethical-culture: exploits cultural and psychological vulnerabilities to manipulate public opinion.
AI-privacy-invasion: synthesizes data to conduct surveillance and compromise privacy and political freedom.

What are the five categories of unethical AI applications identified in the chapter?

Related questions