Artificial intelligence and freedom of expression: implications for democracy

The use of algorithms in digital media to personalise content has important implications for the democratic role of the media. EU-funded research is helping to ensure that these powerful new tools are harnessed for the benefit of society as a whole.

Countries
Countries
  Algeria
  Argentina
  Australia
  Austria
  Bangladesh
  Belarus
  Belgium
  Benin
  Bolivia
  Bosnia and Herzegovina
  Brazil
  Bulgaria
  Burkina Faso
  Cambodia
  Cameroon
  Canada
  Cape Verde
  Chile
  China
  Colombia
  Costa Rica
  Croatia
  Cyprus
  Czechia
  Denmark
  Ecuador
  Egypt
  Estonia
  Ethiopia
  Faroe Islands
  Finland
  France
  French Polynesia
  Georgia

Countries
Countries
  Algeria
  Argentina
  Australia
  Austria
  Bangladesh
  Belarus
  Belgium
  Benin
  Bolivia
  Bosnia and Herzegovina
  Brazil
  Bulgaria
  Burkina Faso
  Cambodia
  Cameroon
  Canada
  Cape Verde
  Chile
  China
  Colombia
  Costa Rica
  Croatia
  Cyprus
  Czechia
  Denmark
  Ecuador
  Egypt
  Estonia
  Ethiopia
  Faroe Islands
  Finland
  France
  French Polynesia
  Georgia


  Infocentre

Published: 9 November 2020  
Related theme(s) and subtheme(s)
Artificial intelligence
Frontier research (ERC)
Information societyInformation technology  |  Internet  |  Multimedia
Research policyHorizon 2020
Science in societyEducation & popular sciences  |  Education & popular sciences  |  Ethics  |  Governance  |  Public opinion  |  Science communication  |  Science communication
Countries involved in the project described in the article
Netherlands
Add to PDF "basket"

Artificial intelligence and freedom of expression: implications for democracy

Image

© Monster Ztudio #236886932 source:stock.adobe.com 2020

Diversity is one of the core principles of democratic participation in society. Thus, new technological innovations which affect the ways in which we receive information online are naturally giving rise to questions and concerns. Like other online services, European newsrooms are using artificial intelligence (AI) tools and data analytics to personalise news to match the interests and previous online behaviour of their readers.

This can be extremely helpful in managing information overload, for example, by providing a more tailored service for both users and advertisers. It does, however, also raise some concerns about the potential of algorithmic bias to exacerbate cultural and social prejudices and create filter bubbles or echo chambers.

The PERSONEWS project, funded by the European Research Council, is one of the first to examine, in depth, the impact of algorithmic news recommendations for both users and news media providers. It is shedding important light on how much we should really be concerned and what safeguards and innovations we can put in place to ensure the ongoing protection of freedom of expression and media diversity.

‘We are at a really transformative moment in time and there is a lot of interest in how AI algorithms could change the media we know. We want to look at how we can best use this new technology to enhance the democratic role of news media rather than undermine it,’ says principal investigator, Professor Natali Helberger of the Institute for Information Law at the University of Amsterdam in the Netherlands.

Maintaining diversity

The PERSONEWS project is combining legal and empirical research to answer many of the questions raised by the use of news recommendation algorithms. Through extensive user surveys and working closely with European newsrooms, technical experts and regulators, the project has been able to develop new tools and guidance. This will help to inform the work of both policymakers, on the one hand, and journalists, editors and quality information providers on the other.

‘There are a lot of challenges for journalists and editors in figuring out how to deal responsibly with the power that algorithms give us. We have learnt that they face the same questions as we researchers and we have been working closely with them to think about what new “smarter” algorithms might look like that are able to generate more diverse recommendations whilst, at the same time, respecting user privacy,’ Helberger explains.

‘We now have the prototype of a tool that helps evaluate the diversity of recommendations, which is in a testing phase in Germany,’ she continues. ‘We have also developed a theory which provides a framework to better assess the opportunities and threats of new recommendation algorithms, which has been well received.’

PERSONEWS is also one of the first projects to look specifically into the legal aspects and fundamental rights questions raised by personalised news and to what extent there are gaps in the legal framework and how they should be dealt with.

Although surveys among user groups revealed a generally high level of acceptance by users, who can see the potential benefits, they also indicated that people want more agency and have concerns around privacy and diversity. The conclusion is that users must be offered clear choice and control in order to maintain trust.

Concerns about the creation of filter bubbles have been directly addressed and demonstrated that this very much depends on how smart the methods being used are.

‘The real issue lies more with the power imbalance between the traditional news media and social media platforms,’ Helberger believes. ‘On social media platforms, algorithms are used to engage people, drive advertising value and keep people on the site. There is a fundamental difference between social media and news media. News media have a democratic role to play and need to reflect on how to use technology to fulfil this mission.’

Overall, PERSONEWS is playing an important role in developing directions for future media law and policy in response to the impact of AI and algorithms on media markets. The research carried out so far has contributed significantly to the debate about the regulation of platforms and has put forward concrete suggestions for future policy. In particular, it highlights the need for public-sector support for innovation in the media sector to create a level playing field and to ensure that true diversity continues to be protected and nurtured.

Project details

  • Project acronym: PERSONEWS
  • Participants: Netherlands (Coordinator)
  • Project N°: 638514
  • Total costs: EUR 1 479 515
  • EU contribution: EUR 1 479 515
  • Duration: August 2015 to May 2021

See also

 

Convert article(s) to PDF

No article selected




loading
Print Version
Share this article
See also
More information about project PERSONEWS