AI: Test what you know AI Quiz: Defining harms from AI uses Activity : Defining AI harms 1 / 8 An AI application can write lyrics and then put them to music before creating beautiful advertisements to attract customers to buy its song on the internet. This AI can spread disinformation This AI can manipulate people’s behaviour This AI can make songwriters, musicians and advertising professionals redundant All of the above None of the above 2 / 8 An AI application provides friendly chats for lonely people. This AI can spread disinformation This AI can manipulate its human companion’s behaviour This AI can collect sensitive data about this person All of the above None of the above 3 / 8 An AI algorithm is used to decide which prisoners should be granted parole. This AI can make wrong decisions based on inadequate or biased data This AI can spread disinformation about prisoners This AI can manipulate the prisoners’ behaviour All of the above None of the above 4 / 8 An AI application can merge two images seamlessly so that it seems that two events happened at the same time and in the same place. This AI can be used to spread disinformation This AI can be used to manipulate someone’s behaviour This AI can suffer network faults and power shortages All of the above None of the above 5 / 8 An AI algorithm constantly collects data on a person’s social media activities to analyse their personality traits and preferences, and then chooses which advertisements to present to them. This AI can spread disinformation This AI can control this person’s behaviour This AI can pass on information about this person to third-parties All of the above None of the above 6 / 8 A General Practitioner connects to an AI device on a cloud to present images of a patient’s eye, taken with a special camera. The AI device analyses the image to give a diagnosis on the presence or absence of Diabetic Retinopathy disease (source: based on information in https://www.fda.gov/news-events/press-announcements/fda-permits-marketing-artificial-intelligence-based-device-detect-certain-diabetes-related-eye#:~:text=The%20U.S.%20Food%20and%20Drug,in%20adults%20who%20have%20diabetes.) This device can spread disinformation This device can control human behaviour The device may suffer network failure or other technical fault to misdiagnose All of the above None of the above [Answer: iii] 7 / 8 An AI driver uses sensors and data about traffic rules and conditions to drive a car without human oversight (Source: Based on information in Abbott 2022, p. 2). This driver can spread disinformation about other cars This driver can control the behaviour of humans sitting in the car This car may suffer network failure or data shortage to cause accidents All of the above None of the above 8 / 8 An AI tool is fed big data on two-dimensional amino acids. It then predicts three-dimensional structure of proteins, having found and analysed patterns in the data at a very fast speed (Source: Based on information in Abbott 2022, p. 2). This tool can spread disinformation This tool can control human behaviour This tool may have analysed only a few patterns, missing some others, giving slightly faulty results All of the above None of the above Your score isThe average score is 44% 0% Restart quiz AI Quiz: AI and theories of IP law 1.2: AI and theories of IP laws This quiz shows four scholarly points of view on granting IP protection to AI. In each case choose which of the 4 IP theoretical frameworks – fairness theory, personality theory, welfare theory and cultural theory – is being invoked. 1 / 4 In a liberal society, IP rights need to be justified before being granted. They can promote innovation but if carried out through politically motivated market interventions, they can also cause harm to society. IP protection for AI outputs, therefore, might be justified in a few cases after consideration but not on a wholesale basis. Fairness theory Personality theory Welfare theory Cultural theory 2 / 4 2. The rapidity and scale of music that can be churned out by AI applications in the ‘style’ of famous artists takes the possibilities for ‘unjust enrichment and free-riding’ off the back of famous artists to a whole new level. Law-makers should protect artists from this violation of their IP (based on O’Connor 2022, p. 80). Fairness theory Personality theory Welfare theory Cultural theory 3 / 4 3. If certain conditions were changed so that AI models became more explainable, IP protection for them would benefit everyone. First, the owner could stop others from making commercial use of them without permission. Second, the public would gain knowledge from the explanation of the disclosed invention (based on Deltorn 2022, p. 111). Fairness theory Personality theory Welfare theory Cultural theory 4 / 4 4. The Berne Convention on copyright refers to human creation. This is not likely to cover creations made by AI authors. Moreover, normally the term of copyright protection is linked to the the life of the author, and IP laws make references to the author’s nationality, which also excludes IP protection for creations made by artificial intelligence (based on Frosio 2022, p. 161). Fairness theory Personality theory Welfare theory Cultural theory Your score isThe average score is 100% 0% Restart quiz