The High Court of Kenya is Reshaping Social Media Governance and AI Accountability

AI Generated.

Meareg 2 others v Meta Platforms INC Amnesty International 6 others (Interested Parties).

Download • 364KB

A recent decision issued by the High Court of Kenya has brought renewed hope that social media platforms can be held accountable for the algorithms they deploy on their social media platforms and the content they allow on their platforms.

In Abraham Meareg, Fisseha Tekle & Katiba Institute v Meta (Petition E541 of 2022), Meta challenged the jurisdiction of the court to hear a case challenging its algorithmic and content moderation practices for causing harm across the African continent. The case gives an example of Ethiopia where Facebook was saturated with hate speech, propaganda for war, and incitement to violence content which catalyzed the civil conflict in Ethiopia. It also gave the example of the Dusit D2 terrorist attack which authorities revealed was organized on the Facebook platform.

From 2019 to 2024, decisions on which content would remain on the Facebook platform were made from Kenya, at Facebook's content moderation center. The question before the court was whether the Kenyan court has jurisdiction to hear a case where a digital decision made in Kenya has an effect both inside and outside Kenya.

Justice Mugambi has now affirmed the jurisdiction of the court to hear human rights violations saying:

'The Petition raises fundamental concerns on acts or omissions that may have been made regarding content posted on Facebook Platform by content moderators based in Kenya that may impact on observance of human rights beyond Kenya through the use of social media. This in my view perfectly falls within the purview of this Court’s jurisdiction to consider under Article 165 (3) (b) of the Constitution.'

The judge also held that the petition raises the following substantial questions of law which deserve determination by a panel of judges:

i. Protection of human rights in the age of Artificial Intelligence (AI), particularly where the use of Artificial Intelligence in social media platforms results in violation of human rights, hence to what extent, if any, social media platforms acting as intermediaries should be held accountable.

ii. Whether there is intentional algorithmic bias and discriminatory AI for different geographical regions.

iii. The Constitutional jurisdiction of this Court in relation to digital decisions made in Kenya with a wider reach on violation of human rights in other jurisdictions and vice-versa.

The matter has now been referred to the Chief Justice for constitution of a bench. Meta sought leave to appeal this decision.