Citizen.digital reports that Kenya and American AI company Anthropic signed a Joint Declaration on artificial intelligence partnership in New York on 22 September 2026, with Foreign Affairs PS Korir Sing’Oei signing for Kenya and Elizabeth Kelly, Head of Beneficial Deployments at Anthropic, signing for the company, the one operative clause being that any future programme involving Kenyan data is governed by Kenyan law.
Goa’s Department of Information Technology, Electronics & Communications, working with Goa Electronics Limited, has launched an AI-powered WhatsApp chatbot for its Goa Online services, on a portal that has received over 21 lakh requests and more than 9 lakh registered users, covering application status, document downloads and payment receipts.
Lydia Manikonda, Mei Si, Sirajam Munira, Oshani Seneviratne and Kristin Bennett report that, in survey responses from 118 students across 12 AI-related courses, academic benefit was associated with early reliance and academic task support, positive impact with cognitive reliance and instructor policy, and negative impact with early reliance and attitudinal change, meaning the same tool produced opposite outcomes depending on how and when it was used.
A Brazilian legal commentary recalls that the Tribunal de Contas da União found in 2017 that federal agencies do not exchange data they already hold and shift the burden of collecting certificates onto citizens, and argues that with gov.br consolidated and AI systems now running inside several agencies, the finding still holds because Brazil’s digitalisation bottleneck was never technological.
Today’s lead is No AI safety index caught a chatbot mistranslating smallpox as syphilis.