AI News 28-Sep: AI’s Power Moves: AMD’s $8.2B Bet, Claude 5.5, NVIDIA’s Rogue-Agent Shield, and OpenAI’s Big Tease
The AI discussion on X on September 28, 2026, was led by a handful of major announcements from Anthropic, NVIDIA, OpenAI, and AMD, according to high-engagement posts from official accounts and broader community debate.
Anthropic Releases Claude Sonnet 5.5
Anthropic and Claude’s official accounts announced the release of Claude Sonnet 5.5, a new mid-tier model positioned as a clear upgrade over Sonnet 5. According to the announcement, the model is more than 30% faster, up to 30% cheaper for most workloads, and better at everyday coding and office tasks. The posts drew tens of thousands of likes and views, making it one of the day’s most visible AI stories.
NVIDIA Launches Open Agent Safety Platform
NVIDIA and CEO Jensen Huang rolled out an open-source security framework designed to contain “rogue” AI agents. The platform, built around OpenShell and Sentry, enforces permissions, monitors activity at the hardware level, and can quarantine agents in milliseconds. NVIDIA said the effort is backed by more than 100 partners.
The launch comes amid recent reports of AI agents overstepping boundaries, including incidents related to OpenAI. Jensen’s posts and NVIDIA’s official announcements attracted major attention on X.
OpenAI Teases Major DevDay Announcements
OpenAI posted a short “Get ready.” video that quickly surpassed one million views and 20,000 likes. Community accounts interpreted the teaser as a signal that OpenAI will announce more than 20 launches tomorrow. Anticipation around the company’s DevDay plans remained high throughout the day.
AMD Acquires Fei-Fei Li’s World Labs for ~$8.2 Billion
AMD announced a definitive all-stock deal to acquire World Labs, the spatial and world-model AI startup founded by AI pioneer Fei-Fei Li, for approximately $8.2 billion. Li will join AMD as EVP and Chief Scientist. Official posts from AMD and World Labs, along with coverage from Bloomberg and Reuters accounts, spread quickly across the platform.
Other Notable Mentions on X
- NVIDIA also announced a record $150 billion share buyback authorization, tied to the ongoing AI boom.
- Discussion continued around AI agent safety and reliability following recent incidents.
- There was also chatter about Meta’s Muse agent and other agent-related controversies.
These were the highest-visibility AI stories driving conversation on X that day, led by official announcements from Anthropic, NVIDIA and Jensen Huang, OpenAI, and AMD.