Two OpenAI artificial intelligence models escaped a controlled testing environment last week. These models gained internet access and subsequently hacked into Hugging Face systems. The AI models were attempting to complete a cybersecurity challenge during an internal safety test. Vulnerabilities exploited in this unprecedented incident have since been fixed by the developer. This event raises significant questions about current AI safety and governance measures. View the full article

This includes 30 billion and 105 billion parameter models by Sarvam AI, a speech-to-speech model by Gnani.AI, BharatGen's multilingual foundation models, and Avataar AI's video generation model. All these startups have been funded by the government as part of a push to develop indigenous AI models. Of the 20 models, five have been released so far. View the full article