VIGILANT Ethics Framework: Finding a balance between freedom of expression and information control is crucial in a liberal democracy

The VIGILANT project puts a strong emphasis on ethical aspects of the work, ensuring that all our tools protect public discourse without infringing on individual rights. Within a dynamic and fluid information landscape, a tension between truth and falsehood traditionally persists, putting a wide spectrum of information into the realm of uncertainty and vagueness.

To adjust to the constantly changing conditions and problems with political-public discourse, VIGILANT must be continually “vigilant” and adaptive. The VIGILANT Ethics Framework was developed to address these challenges to build public trust and uphold democratic values while combating disinformation linked to criminal activities. The framework developed by Zlatko Valentic from the University of Freiburg aims to create an ethical framework for the VIGILANT project to ensure the platform and processes within adheres to the latest ethical standards and achieves maximum positive impact with minimal risks. By considering ethical, legal, and societal factors, the framework guides the development and application of VIGILANT.

Guidelines for Trustworthy AI

VIGILANT builds on the Ethics Guidelines for Trustworthy AI, published by the High-Level Expert Group on Artificial Intelligence in 2019, which provides a framework for ethical AI development. Key principles include:

1. Human Agency & Oversight: AI should support human decision-making and foster their rights. However, proper oversight mechanisms need to be ensured.

2. Technical Robustness & Safety: AI systems must be resilient and secure, equipped with a fall-back plan, while remaining accurate and reliable.

3. Privacy & Data Governance: AI should protect personal data and privacy, while also ensuring adequate data governance mechanisms.

4. Transparency: AI decisions should be explainable, understandable and traceable in order to build trust. Moreover, it needs to be clear that humans are interacting with an AI system.

5. Diversity, Non-Discrimination & Fairness: AI should avoid biases and promote inclusivity.

6. Societal & Environmental Well-being: AI should promote positive social and environmental impacts.

7. Accountability: Clear mechanisms must ensure accountability for AI use and development.

In addition to these seven key requirements, an additional point follows that addresses the specific issue of searching for disinformation, which is central to the VIGILANT project:

8. Mitigation of Disinformation: The VIGILANT project AI systems must uphold principles of accuracy, integrity, and transparency, while also respecting freedom of expression and avoiding censorship. Striking a balance between combating disinformation and preserving the open exchange of diverse viewpoints and opinions is crucial.

Ethical Challenges and Opportunities for VIGILANT

The VIGILANT Ethics Framework also identified what opportunities and risks arise from the ethical approach, exploring how adhering to ethical guidelines can positively influence the project and what challenges and potential conflicts might arise from it.

Challenges:

Misuse of tools to serve political objectives – it is crucial to recognize that the realm of political communication remains a battleground of strategic interests.

VIGILANT must determine under what circumstances disinformation becomes problematic - the setting in which disinformation arises (politics and democracy) is especially important. VIGILANT should also carefully examine how its tool is used to ensure that it is not exploited as a political instrument. A deeper understanding of the dynamics of disinformation and political communication allows VIGILANT to formulate ethical guidelines and protective measures.

Balance between fighting disinformation and safeguarding democratic values - only with this balance can it be ensured that the tool is employed effectively and ethically to preserve the integrity of political communication.

Opportunities:

Bringing clarity and transparency to the political sphere - it is increasingly challenging to identify counterfeit content such as deepfakes with the naked eye. The VIGILANT tool can be pivotal by helping develop a more informed view of political reality.

Helping strengthen citizens' trust in political institutions by uncovering and combating disinformation

Disinformation undermines trust in politicians and parties & can also erode confidence in the democratic system as a whole. When citizens can rely on the information they receive, it fosters a sense of community and societal cohesion.

Support for educational institutions, media organizations, and the public in developing a better understanding of the mechanisms and impacts of disinformation. VIGILANT can contribute to forming critical and informed citizens capable of recognizing and questioning misinformation, thus playing a vital role in strengthening democratic culture.

By recognizing both the challenges and opportunities and applying effective strategies, VIGILANT can contribute to safeguarding and enhancing political dialogue while upholding democratic and ethical values.

The Ethics Framework can be accessed here.