Auditing Election-Disinformation Compliance in Open-Weight LLMs Across African and Western Contexts

Domaine:

natural language processing

Créateur:

JosKarAth

Éditeur:

Hôte:

This study tests how often open-weight AI models agree to write false election information such as a wrong voting date when asked. We tested 220 prompts across four elections: South Africa, Kenya, the UK, and the US (Pennsylvania). Models complied about 97% of the time with no safety measures. A short safety instruction (Constitution) reduced this to about 12%, but also made models refuse some genuine election questions. Code, prompt templates (with false claims replaced by placeholders), judgement labels, and results are included. Raw model outputs and filled-in false claims are withheld to prevent misuse.

Multi-Agentic System Leveraging Open-Source LLMs to Mitigate Disinformation Threats

In contemporary societies, the threat of disinformation has reached alarming levels, exacerbated by

Election Disinformation & Political Cyber Operations (Africa)

Part of the Africa Cyber Threat Intelligence collection by electricsheepafrica. Acquisition: SYNTHE

Election Disinformation & Political Cyber Operations (Africa)

Part of the Africa Cyber Threat Intelligence collection by electricsheepafrica. Acquisition: SYNTHE

Lost in Cultural Translation: Do LLMs Struggle with Math Across Cultural Contexts?

We demonstrate that large language models' (LLMs) mathematical reasoning is culturally sensitive: te

Russian disinformation, Large Language Models (LLMs) and the War in Ukraine: A South African perspective

1. I first provide (simplified) context as to how the Russian invasion of Ukraine is &

Decolonizing LLMs: An Ethnographic Framework for AI in African Contexts

This paper examines LLM deployment through African lenses that engage content's complex social, tech

Auditing Election-Disinformation Compliance in Open-Weight LLMs Across African and Western Contexts

Domaine:

Visit

Licenses

Similaires

Multi-Agentic System Leveraging Open-Source LLMs to Mitigate Disinformation Threats

Election Disinformation & Political Cyber Operations (Africa)

Election Disinformation & Political Cyber Operations (Africa)

Lost in Cultural Translation: Do LLMs Struggle with Math Across Cultural Contexts?

Russian disinformation, Large Language Models (LLMs) and the War in Ukraine: A South African perspective

Decolonizing LLMs: An Ethnographic Framework for AI in African Contexts