Code of Practice on Disinformation: new reports available in the Transparency Centre

This is the first reporting round covering a full, 6-month period.

Four new signatories of the Code (Alliance4Europe, Newtral, EFCSN and Seznam) have submitted their baseline report at the same time, in accordance with their commitments under the Code.

The reports can be accessed through the Transparency Centre of the Code, established and maintained by the signatories of the Code.

Signatories who are major online platforms committed to report every six months on their actions taken under the Code, while other signatories report once per year.

The platforms’ new reports

The new reports follow the harmonised reporting template consisting of 152 reporting elements (111 qualitative reporting elements, and 42 service level indicators/quantitative indicators) across the Code's chapters.

Overall, this new reporting round attests of the signatories’ efforts to provide further insight into their actions to fight disinformation, with more stable data covering a full 6-month reporting period. Platforms generally make improvements in providing more granular and insightful data, closing some data gaps identified in their baseline reports (from February 2023). At the same time, further efforts are needed to provide more targeted, complete and meaningful data.

Given the potential of generative AI for creating and disseminating disinformation, platforms also report about their recent efforts to provide safeguards regarding new generative AI systems on their services.

The reports also include insights on platforms’ actions to reduce disinformation surrounding Russia’s war in Ukraine. In view of the important risks posed by disinformation to our democracies the Commission expects signatories to continue their work and increase their efforts to fight disinformation on Ukraine and around elections.

The next set of reports are due in early 2024 (with information and data covering the second half of 2023) and should provide further insights on the Code’s implementation. Those reports will also contain information on how signatories are preparing and putting in place measures to reduce the spread of disinformation ahead of the 2024 EU elections, while continuing to report on their efforts in the context of the war in Ukraine.

A first set of Structural Indicators

For the first time, the reporting exercise also includes the publication of an initial set of Structural Indicators, providing insight about disinformation on platforms’ services.

Signatories selected an independent third party, TrustLab, to define and implement a methodology to measure two Structural Indicators for a first, pilot phase: Prevalence and Sources of disinformation. Those have been measured in an initial set of three Member States (Poland, Slovakia, Spain). Trustlab’s study covers all platform signatories to the Code at the time of the launch, including Twitter (now X) and is available on the Transparency Centre, along with the platforms’ reports.

The Commission expects signatories to continue their work on the structural indicators by expanding and fine-tuning them as well as include additional Member States in the measurement. Signatories should also develop additional structural indicators to measure other important characteristics of the disinformation phenomenon, such as the monetisation of disinformation

Some examples of data from the reports

• Google indicates that in the first half of 2023 it prevented more than EUR 31 million in advertising from flowing to disinformation actors in the EU. Also, Google served 20 441 political ads in the EU for a value of almost EUR 4.5 million, while rejecting 141 823 political ads for failing the identity verification processes.
• Meta reported that over 40 million pieces of content received a factchecking label on Facebook and over 1.1 million on Instagram. 95% of users encountering content with a warning label that the information has been fact-checked choose not to click on it, and 37% of users on Facebook and 38% on Instagram who intended to share fact-checked content opt to cancel their sharing action when receiving the warning.
• TikTok reported that 29.93% of users cancel their sharing action when encountering an 'unverified content’ label on the platform. Also, 140 635 videos accounting for more than 1 billion views were removed from the platform for infringement of the misinformation policy.
• Microsoft reports that more than 6.7 million fake LinkedIn accounts were prevented from being created (blocking registration) or restricted in the EU for the first half of 2023.

Additional data examples from the section on platforms’ efforts to counter disinformation on Russia’s war of aggression in Ukraine:

• Google reports that, between January and April 2023, YouTube terminated 411 channels and 10 Blogger blogs involved in coordinated influence operations linked to the Russian-state sponsored Internet Research Agency (IRA).
• Meta reports expanding its factchecking partnerships to 26 partners covering 22 languages in the EU, which now includes also Czech and Slovak.
• TikTok’s fact-checking efforts cover Russian, Ukrainian, Belarusian and 17 European languages, including through a new partnership with Reuters. In this context, 832 videos related to the war have been fact-checked, of which 211 have been removed as a consequence of fact-checking.
• Microsoft highlights that Bing Search has either promoted authoritative information or downgraded less authoritative information in relation to 798 180 search queries related to the crisis

More information on the Code of Practice on Disinformation