- Google has updated its Gemini AI system with a new, larger model named Argon in a bid to compete with rival frontier models from OpenAI and Anthropic.
- The tech giant is restricting initial access to select cybersecurity partners and the Trump administration's voluntary pre-release process due to concerns over potential cyber security threats.
- Although Argon surpassed competing models like OpenAI's Astra and Anthropic's Opus on several industry benchmarks, it still trails behind on specific coding-related metrics.
- Following leadership departures and an overhaul at its DeepMind AI lab, Google confirmed it has scrapped plans to release Gemini 3.5 Pro, which was originally scheduled for a June launch.
- Months of delays have heightened accusations that Google is losing ground to rivals, prompting chief executive Sundar Pichai and the company to pivot towards promoting Gemini's cost advantages.
IN FULL
