author

Moral & Legal Challenges Presented by Artificial Intelligence

  • Programmed versus Interpretive Bias: A critical challenge can arise based on the balance between how machines are programmed to act (reflecting our value sets) versus their own judgement
  • This performance has a large bearing on how much accountability and trust can be placed on AI-based results
  • Below are some questions from different industries where AI is expected to bear a large impact and drive autonomous work:
  • - If a self-driving car faces an inevitable situation of a crash, how does it decide what or who to hit?
  • - If a self-driving car is deemed to be at fault of getting hit, who faces the ultimate responsibility: (i) the car owner, (ii) the car manufacturer, or (iii) the AI developer ?
  • - If an algorithm is tasked with predicting the likelihood of someone (either incarcerated or otherwise) committing a crime, what parameters does it use ?
  • - If an algorithm detects probability of a crime being committed, what preventive actions can be taken and are those morally correct because crime hasn't been committed and it's only a partial probability ?
  • - If a social media platform such as Facebook makes an algorithm to detect Fake News, what parameters does it use and how much accuracy is considered acceptable ?
  • - If content shared with people after filtering fake news using AI is deemed as misleading or fabricated, will it be considered intentional or by mistake, and who bears the responsibility for the same ?
  • These above questions are very real scenarios in eliminating the sociological bias from algorithms, especially when they are designed to mimick human judgement
  • Moreover, the modern Anglo-Saxon judgement system is based on accountability of specific individuals or entities such as companies and governments, and isn't trained to consider factors involved in AI performance
  • Digital technology has advanced 30 years faster than any regulatory systems to constrain adverse consequences, and identifying mechanisms to enforce accountability on inanimate machines
  • AI is advancing faster than regulations and laws can adapt, putting the debate into esoteric realms of moral philosophy

Don't forget to share this pointer!

View more comments +