Full Summary
This Thursday morning, OpenAI has launched its new AI model, GPT-6 Astra, with both Geo News and Gizmodo confirming the company's claim that it represents a significant leap forward, potentially signaling the start of the artificial general intelligence era. Multiple sources, including OpenAI itself and Mashable, confirm Astra is rolling out today to select organizations, and will soon be available to ChatGPT Plus, Pro, Business, and Enterprise users, as well as via the OpenAI API and Amazon Web Services. Geo News reports OpenAI President Greg Brockman suggests Astra could signal the start of the artificial general intelligence era, where AI performs most economically valuable work at a human level. TradingKey adds that Astra is priced at about 2.5 times its predecessor, and can autonomously complete longer, more complex tasks, like compressing a 30-minute human search to just over 5 minutes. Here's the thing: Astra introduces significant cybersecurity risks. TradingKey, Geo News, and Mashable all highlight Astra's ability to identify unknown system flaws and create exploits autonomously. Mashable notes Astra is the first OpenAI model designated a "critical" cybersecurity risk, with the company confirming this risk designation the day before launch. Despite this, OpenAI moved forward, implementing strict enterprise rollout restrictions and enhanced security safeguards. Meanwhile, Meta has launched Muse Spark 1.3, which SiliconANGLE and Livemint confirm is now available to developers and will soon roll out to Facebook, Instagram, and the Meta AI application. Both SiliconANGLE and Livemint report Meta's Chief AI Officer, Alexandr Wang, claims Muse Spark 1.3 is competitive with Anthropic's Claude Fable 5.1 and better than OpenAI's GPT-5.6 Sol, especially in coding. BeInCrypto adds that Muse Spark 1.3 excels in agentic knowledge work and scientific reasoning. What nobody expected: Inshorts reports Wang openly questioned Google's Gemini AI model, stating, "I really hate to say...but…gemini who?" Google also launched Gemini 3.8 Flash, along with a cybersecurity version, as confirmed by BeInCrypto and Fortune. BeInCrypto indicates Gemini 3.8 Flash has an advantage in factual recall and terminal coding, with the cybersecurity variant scoring 86.2% on a benchmark for finding vulnerabilities. This intense competition means your next digital interaction could be with an AI far more capable, but also potentially more risky, than anything you've seen before.