
Following OpenAI, Google Acknowledges Safety Setbacks in New AI Model
The newer model underperforms on two key automated safety benchmarks: "text-to-text safety" and "image-to-text safety"
The newer model underperforms on two key automated safety benchmarks: "text-to-text safety" and "image-to-text safety"
Meta AI offers more personalised responses by tapping into information users have already shared, such as their profiles, preferences, and online interactions
Recently, Microsoft CEO Satya Nadella told Zuckerberg at Llama Con that AI generates between 20% and 30% of the code in the company’s repositories
His bold predictions often don't materialise
The rollout is global, reflecting a broader shift where AI tools are integrated into users’ everyday digital spaces
OpenAI stated that its policies prohibit such content for users under 18 and acknowledged the issue was due to a bug
In a significant recognition for Meghalaya, a team of m Master’s candidates from the Harvard Kennedy School has secured first place at the prestigious “AI for Good Hackathon,” organised by the Centre for International Development at Harvard University, the Shillong Times reported. Their winning project, MEGHA — short for Meghalaya
Space Llama evolves and builds upon successful efforts by Booz Allen to deploy and operate the first known large language model (LLM) in space.
Private AI investment in the U.S. soared to $109.1 bn—outpacing China’s $9.3 bn
India’s first dedicated AI Lab for Railway Safety was inaugurated on April 24, 2025, by Sri Praveen Kumar IRSE, Managing Director of DFCCIL, in the presence of IISc faculty and representatives from a railway-focused IISc startup. Located at the Research Centre (iHUB) within the IISc campus, the lab aims
Firefly, which Adobe introduced in 2023, has until now featured the company’s own AI models
AI Startups
It works on older devices like the iPhone 13 mini, unlike Apple Intelligence