The European Commission wants input on AI misinformation for election guidelines
The European Commission wants tech platforms like TikTok, X and Facebook to protect the upcoming European elections from disinformation.
The Commission has launched a public consultation on the proposed Election Security Guidelines for Very Large Online Platforms (VLOPs) and Very Large Online Search Engines (VLOSEs). The recommendations aim to reduce the risks to democracy through generative AI and deep fakes.
The draft guidance provides examples of measures to address election-related risks, including specific measures related to generative AI content, planning to mitigate risk before or after an election event, and providing clear guidance for European Parliament elections.
Generative AI can mislead voters or manipulate electoral processes by creating and distributing inaccurate, synthetic content about political actors, false descriptions of events, polls, contexts or narratives.
Ensuring electoral integrity means protecting public discourse – much of it online.
We are seeking views on our Digital Services Act guidance to protect the building blocks of our democracy to address emerging online threats to electoral processes ↓
— European Commission (@EU_Commission) February 8, 2024
Currently open for public consultation in the EU until March 7, the Election Security Directive advises users to notify users of potential errors in content generated by generative AI in appropriate platforms.
According to the draft, the guidelines will point users to authoritative sources of information and state that tech giants must implement safeguards to prevent misleading content that has a significant impact on user behavior.
In the AI-derived text, the current recommendation for VLOPs/VLOSEs is to “indicate, where possible, the presence of objective sources used as input data to enable users to verify reliability and further contextualize the data.
The draft directive draws inspiration from the EU's recently adopted legislative proposal, the AI Act, and the non-binding AI Treaty on “best practices” for risk prevention.
Related: EU Artificial Intelligence Law Explained
Concerns about advanced AI systems such as large language models have led to widespread recognition of tools such as OpenAI's ChatGPT after the 2023 Generative AI viral boom.
The Commission did not say exactly when companies would be required to label content processed under the EU's content moderation law under the Digital Services Act.
However, in a company blog post, Meta announced plans to introduce new guidelines on AI-generated content on Facebook, Instagram and Threads in the coming months. Any AI-generated content, whether through metadata or intentional watermarking, receives a visible tag.
Magazine: AI killed the industry': EasyTranslate boss adapting to change
