Individuals today live in the world of Artificial Intelligence (AI). It permeates our lives in numerous ways, either directly or indirectly, by performing tasks that, until quite recently, could only be performed by a human with specialized knowledge and training. Some of the examples of the use of AI in our daily lives include spam filters, Alexa, Siri, driverless cars, automatic vacuum cleaners etc. These technologies are used in different sectors like technological industries, healthcare, education, transportation, defense, law and agriculture, among many others. This list keeps on increasing with time owing to the technological advancements taking place in the modern world.
The increasing ubiquity and rapidly expanding commercial potential of AI in different sectors has spurred massive private sector investment in the AI projects. The potential for further rapid advances in AI technology has prompted expressions of alarm, wherein some scholars highlight the role of government in regulating the development of AI and putting restrictions on AI operations.
Learn more about Artificial Intelligence with Enhelion’s Law Firm certified Online Course!
Before understanding what steps have been taken by different jurisdictions to address the implications of growing use of AI, it is important to understand exactly what issues relating to AI needs government deliberation. Broadly, the most significant issues can be listed as- data protection concerns, the issue of biasness, the challenges to the Intellectual Property regime, the issue of civil liability as well as the impact of AI in the criminal justice system.
As AI continues to find its way into our daily lives, its propensity to interfere with human rights only gets more severe. Privacy is recognized as an international human right that is essential to human dignity and is inalienable. Data protection also forms an intrinsic part of an individual’s privacy as privacy can be guaranteed only when the personal and sensitive personal data of an individual is protected. Taking this implication into consideration, many jurisdictions now recognize the right to data protection, as part of right to privacy, and therefore, a fundamental right. The increasing use of AI creates an issue in the domain of data protection as the AI systems are often trained through access to and analysis of big data sets. This makes the right to data protection susceptible to violation as there is a possibility of leak of personal data or sensitive personal data of an individual by an AI.
Machine learning is a type of artificial intelligence which uses data sets to understand the pattern and learns to perform a particular task from it. Because machine learning algorithms use data sets to derive a conclusion, if the data set provided to the machine learning system is itself biased, the system will provide biased results. Furthermore, since AI is a technology, it is susceptible to minute measurement errors. However, such minute errors might lead to a big problem with millions of users, when thousands could be affected by error rates. These two issues create biased results. For example, Google Photos’ image recognition software, in 2015, labelled photos of black people as ‘gorillas’. Owing to the complex software model, the developers were unable to resolve the issue. Therefore, the only recourse available was to remove any monkey-related words from the data which was fed to the AI.[1]
Learn more about Artificial Intelligence with Enhelion’s Law Firm certified Online Course!
With respect to the impact of AI in the intellectual property (IP) regime, there are two sets of implications. Firstly, the impact of law on the creations of artificial intelligence, and secondly, the impact of AI on law i.e. how AI technology is being used in the field of law. With respect to the former, there is much debate going on around the world with respect to giving the AI the IP rights of its own creations, without human intervention. Since some types of complex AI can ‘invent’ or ‘create’ creative work without human involvement, there is a unanimity in addressing the legal status of such invention or creation. This is where the role of patent and copyright regime comes into picture. Although a number of applications mentioning AI as an inventor, have been filed in the patent offices of different jurisdictions, however, these jurisdictions have shown their reluctance in granting the AI intellectual property rights. In the similar way, courts in various jurisdictions have reiterated that copyright can only be granted for works which involve human as a creator. However, it should be kept in mind that AI technology is developing and in future, there might be cases where AI invents or creates something which the programmer of such AI has no idea about, or has no role to play. In such a situation, is it justified to not develop the IP regime to accommodate such inventions and creations.
There are many documented cases of AI gone wrong in the criminal justice system. Machine learning is often used for risk scoring the defendants as a means to remove human bias prevalent in judges while adjudicating cases with respect to sentencing and bail. Predictive policing, on the other hand, is used as a measure to allocate police resources to prevent crime. However, the conclusions arrived at using such AI systems in the criminal justice process re-introduces the element of bias against the accused, which these AI systems were aimed at addressing. Another issue with respect to the intersection of AI and criminal law is the ambiguity with respect to the criminal liability in case AI does an act, which, if done by a human, would constitute an offence under the criminal law.
Lastly, the issue of civil liability of AI highlights the situation where the traditional concepts of conceptualizing liability is being adopted by the courts if a product liability claim involves an AI.
Learn more about Artificial Intelligence with Enhelion’s Law Firm certified Online Course!
[1] WIRED, https://www.wired.com/story/when-it-comes-to-gorillas- google-photos-remains-blind/ (last visited Feb. 1, 2021).