AI News Roundup — April 1, 2026
Anthropic enters early IPO discussions targeting Q4 2026, Disney was blindsided by OpenAI's Sora shutdown, Alibaba releases a full omnimodal model, and Apple cracks down on vibe-coding apps.
Anthropic in Early IPO Discussions Targeting Q4 2026
Anthropic is in early-stage IPO discussions targeting as early as October 2026, with Goldman Sachs, JPMorgan Chase, and Morgan Stanley competing for underwriting roles. The company closed a $30 billion round in February at a $380 billion valuation and is approaching $19 billion in annualized revenue.
Disney Blindsided by OpenAI's Sora Shutdown
Disney, which committed $1 billion to an OpenAI partnership, reportedly learned of the Sora shutdown less than an hour before the public announcement. Sora's global users had peaked around one million then collapsed below 500,000 while operational costs held steady at roughly $1 million per day.
Alibaba Releases Qwen3.5-Omni Full Multimodal Model
Alibaba released Qwen3.5-Omni, a full omnimodal large language model that understands text, images, audio, and video in a single system with real-time capabilities. The model advances China's position in the multimodal AI race, though enterprise adoption remains limited.
Apple Cracks Down on Vibe-Coding Apps
Apple removed the app Anything from the App Store for violating its rules, escalating a crackdown on vibe-coding software. AI coding tools are spreading faster than platform policy can adapt, and if Apple takes a harder line, it could affect startups trying to lower the barrier to software creation.
California Governor Signs AI Regulation Executive Order
Governor Gavin Newsom signed an executive order giving California four months to develop AI policies focused on public safety, covering harmful bias, unlawful discrimination, surveillance, and watermarking of AI-generated media. The order signals the most aggressive state-level AI regulation push in the US.
Mirror Test Reveals Claude Opus 4.6 Shows Self-Recognition
A proposed Mirror Test for LLMs found that Anthropic's Opus 4.6 shows notable self-recognition capabilities, outperforming OpenAI's GPT models at identifying its own outputs without explicit cues. However, no LLM demonstrated consistent self-awareness across all test conditions.
Reddit Requires Human Verification for Suspicious Accounts
Reddit became the first major platform to draw a line between AI-assisted human and fully automated bot, requiring human verification for suspicious accounts as AI-generated content floods social platforms. The move could set a precedent for other platforms grappling with bot traffic.