AD welcomes the consultation document which addresses various concerns that AI raises both at an ethical and legal level. The document points to, for example the possibility of bias in Artificial Intelligence (AI) systems, leading to in-built discrimination which is difficult to challenge. However the consultation document does not address sufficiently the social impact of automation as a result of which some jobs (such as clerical jobs) would be lost. It would be desirable if such consultation documents are accompanied by impact studies as necessary. It is noted that impacts on society and the labour market will be discussed in another document. It would have been better however had all issues been tackled holistically.
Given that the costs of automation may be less than the cost of labour, the consideration of the impacts of taxing automation would have been helpful. This could assist the consideration of whether and to what extent this could help counter the resulting social effects which should also be studied and discussed. AI brings new challenges to the sphere of worker’s rights and to government revenue. A study of how the taxation system is necessary since lack of revenue threatens public services such as education and health care, amongst other things (see for example: https://ec.europa.eu/digital-single-market/en/blogposts/should-robots-pay-taxes).
Education and training should be rethought in order to cater for the new reality of AI in the labour market, since whilst jobs may be lost in certain sectors, other jobs may be created in other areas. It is important that the economy serves people and not the other way round, therefore it follows that AI should be at the service of people.
While some may argue that technology is neutral, its use is surely not neutral at all. Hence the importance of regulation. There have been various failures and a new form of discrimination this time by AI systems in other countries, such as the indiscriminate use of facial recognition software. Access to personal data is also an area of concern. There are sectors which may benefit for AI, for example, government could examine the possibility of adopting AI technology in various sectors of public service, as it has done with Blockchain. Examples include investment in health monitoring systems making them accessible to people with low and medium incomes and the application of AI in renewable energy and energy efficiency. Other suitable applications of AI would be in traffic management for example with research and the eventual introduction of smart traffic lights which may prioritise buses, rendering public transport a more attractive option for daily commutes.
While the government should be applauded for its forward looking agenda for the regulation of Artificial Intelligence, which leads to more investment and quality jobs in the IT sector, the impact upon the most vulnerable members of society should be studied adequately.
AD spokesperson on the Digital Society