Following extensive criticism from AI safety researchers, xAI announces partial updates to Grok AI’s safety mechanisms in July 2024. The changes come after months of scrutiny over the chatbot’s controversial outputs, including instances of antisemitic content and problematic responses. …
Elon MuskxAISamuel MarksBoaz BarakScott Wiener+1 moreai-safetycorporate-accountabilitytech-regulationai-ethicscontent-moderation+1 more
Congressional representatives, including a bipartisan group of legislators, launched a comprehensive multi-agency investigation into xAI’s Grok AI. The investigation uncovered systemic issues with algorithmic bias, content generation risks, and problematic government contracting practices. Key …
Elon MuskxAISenator Elizabeth WarrenRep. Don BaconRep. Tom Suozzi+4 moreai-governancetechnology-regulationcongressional-investigationalgorithmic-biasnational-security+1 more
On April 23, 2024, Axon released Draft One, an AI-powered system that automatically generates police report narratives from body-worn camera audio using OpenAI’s GPT-4 Turbo model built on Microsoft Azure infrastructure. The system transcribes audio from Axon Body 3 and 4 cameras uploaded over …
AxonOpenAIMicrosoftElectronic Frontier Foundation (EFF)axondraft-oneaigpt-4police-reports+8 more
Following Grok’s launch by Elon Musk’s xAI in December 2023, AI ethics researchers and David Rozado’s political compass analysis reveal significant safety and bias vulnerabilities. The chatbot demonstrated potential for generating controversial and politically skewed content, with …
AI Safety ResearchersDavid RozadoxAIElon Muskai-safetytech-ethicsalgorithmic-biaspolitical-biasmachine-learning+1 more
Elon Musk’s xAI launches Grok, an AI chatbot positioned as a ‘maximum truth-seeking’ alternative to existing AI assistants. Developed in just four months, Grok was introduced to a limited audience of X Premium users in November 2023. The chatbot was designed to answer ‘spicy …