African Trust & Safety LLM Benchmark: Stress-testing AI Safety Across Africa’s Languages and Contexts

Supported by the GSMA, the African Trust & Safety LLM Challenge has produced a new benchmark of 4,216 verified and reproducible AI safety stress tests across African languages, multilingual prompts, and code-switched contexts. The dataset was generated through the Zindi community, with 320 participants submitting more than 42,000 adversarial attacks across 4,010 markdown files. After […] The post African Trust & Safety LLM Benchmark: Stress-testing AI Safety Across Africa’s Languages and Contexts appeared first on Newsroom .