Equality commissioner spending €700,000 to study effects of AI

The Office of the Gender Equality and Equal Treatment Commissioner, together with partners from the Ministry of Justice, the Ministry of Economic Affairs, Tallinn University of Technology and the Office of the Equal Opportunities Ombudsperson of Lithuania, will spend €689,571 over 24 months to study the impact of artificial intelligence on equality.
The Office of the Gender Equality and Equal Treatment Commissioner announced that the Equitech project (Improving response to risks of discrimination, bias and intolerance in automated decision-making systems to promote equality) aims to assess how well the creation of automated decision-making systems in Estonia and Lithuania considers the need to ensure gender equality, equal treatment of vulnerable groups and protection of fundamental rights, as well as to prevent discrimination in AI-based public services.
Estonia's Gender Equality Commissioner Christian Veske stated through a press release that his goal is to better understand the mechanisms by which algorithms may result in discriminatory outcomes and how to avoid this.
"To this end, we have combined expertise in equal treatment, legal and IT fields to enhance the capacity of Estonian and Lithuanian public institutions and their partners in addressing the risks of discrimination and bias in AI-created systems," Veske added.
Currently, there are about 140 AI solutions in use in Estonia's public sector, and this number is on the rise. These solutions are applied in a wide range of areas, such as mass digitization of photographic heritage, calculating forest heights, detecting VAT fraud and assisting citizens through virtual assistants in various institutions.
"We must ensure transparency in the processes and the right to appeal decisions for individuals interacting with AI. These directions are also reflected in the AI action plan for 2024-2026, which we aim to soon have approved by the government," explained Ott Velsberg, chief data officer at the Ministry of Economic Affairs and Communications.
The first phase of the project involves creating a methodology based on scientific literature, policy documents and legislation to identify discrimination risks in algorithm-based decision-making systems. Additionally, existing algorithm-based systems in Estonia will be mapped out, and three systems will be tested to see how well they consider potential risks.
The project will result in a universal toolkit, usable across the European Union, for public institutions and IT system developers. This toolkit will include guidelines, a checklist, and e-training materials, according to the office.
Plans include addressing the topic in an international research seminar and scientific article, conducting training sessions for public officials and developers, hosting an international conference, and launching a campaign aimed at the broader public.
--
Follow ERR News on Facebook and Twitter and never miss an update!
Editor: Marcus Turovski