· 001 · blog · 5 min read
OpenAI Acquires TBPN · Microsoft 3 New Models · Google Vids
AI News Briefing
Date: April 3, 2026
Time: 12:00 PM (Asia/Shanghai)
Edition: Daily Briefing #2026-04-03-12
Top Stories
1. OpenAI Acquires TBPN, the Buzzy Founder-Led Business Talk Show
Source: TechCrunch | Time: 6 hours ago
OpenAI has acquired TBPN (The Business Podcast Network), a popular founder-led business talk show that has gained cult status in Silicon Valley. The podcast will continue to operate independently under the oversight of Chris Lehane, who serves as OpenAI’s chief political operative.
The acquisition signals OpenAI’s expanding media strategy as the company prepares for its anticipated IPO. TBPN has become a key platform for tech founders and investors to discuss industry trends, making it a valuable asset for OpenAI’s communications ecosystem.
2. Microsoft Takes on AI Rivals with Three New Foundational Models
Source: TechCrunch | Time: 9 hours ago
Microsoft’s MAI (Microsoft AI) group has released three new foundational models capable of transcribing voice to text, generating audio, and creating images. The models were developed following the group’s formation six months ago and represent Microsoft’s push to compete directly with OpenAI, Google, and Anthropic in the foundational model space.
The new models will be integrated across Microsoft’s enterprise products, including Teams, Office 365, and Azure AI services, giving the company a more comprehensive AI stack for business customers.
3. Google Now Lets You Direct Avatars Through Prompts in Its Vids App
Source: TechCrunch | Time: 10 hours ago
Google has added a new feature to its Vids app that allows users to customize and instruct avatars for video creation through natural language prompts. The update expands Google’s AI-powered video editing capabilities, enabling users to create personalized video content with AI-generated presenters.
The feature positions Google to compete with emerging AI video platforms while leveraging its existing Workspace integration for enterprise adoption.
4. Anthropic Accidentally Takes Down 8,100 GitHub Repos in Botched Takedown
Source: TechCrunch | Time: 18 hours ago
Anthropic caused widespread disruption on GitHub after accidentally leaking source code for its Claude Code command-line application. When attempting to remove the leaked code via DMCA takedown notices, the company’s action affected approximately 8,100 repositories—including legitimate forks of Anthropic’s own public Claude Code repository.
Boris Cherny, Anthropic’s head of Claude Code, acknowledged the error and retracted most takedown notices, limiting enforcement to a single repository and 96 forks containing the accidentally released source code. GitHub has since restored access to the affected forks.
The incident raises questions about Anthropic’s operational readiness as the company reportedly plans an IPO, with legal experts suggesting potential shareholder lawsuits over the source code leak.
5. Meta’s Hyperion AI Data Center Will Be Powered by 10 New Natural Gas Plants
Source: TechCrunch | Time: 22 hours ago
Meta’s upcoming Hyperion AI data center will be powered by 10 new natural gas plants, with energy capacity equivalent to powering South Dakota. The massive infrastructure investment underscores the enormous power requirements of training and running large-scale AI models.
The announcement has drawn criticism from environmental groups concerned about the carbon footprint of AI infrastructure expansion, even as Meta commits to broader sustainability goals.
6. Cognichip Wants AI to Design the Chips That Power AI, Just Raised $60M
Source: TechCrunch | Time: 24 hours ago
Cognichip has secured $60 million in funding to develop AI systems that autonomously design the chips powering AI workloads. The startup claims it can reduce chip development costs by more than 75% and cut the timeline by more than half.
The funding round reflects growing investor confidence in AI-native hardware design, as traditional chip development cycles struggle to keep pace with rapidly evolving AI model requirements. The company aims to close the loop between AI software and hardware by using machine learning to optimize chip architectures specifically for neural network operations.
7. Mercor Says It Was Hit by Cyberattack Tied to LiteLLM Compromise
Source: TechCrunch | Time: 2 days ago
AI recruiting platform Mercor disclosed a cybersecurity breach tied to the recent compromise of the open-source LiteLLM project. The attack underscores cascading risks in the AI supply chain, where vulnerabilities in widely-used open-source components can propagate across multiple downstream services.
The incident has renewed calls for enhanced security auditing of AI infrastructure dependencies and more robust supply chain security practices.
Trend Watch
📈 AI Media Consolidation: OpenAI’s acquisition of TBPN signals a new phase in AI company media strategies. As AI labs mature and approach public markets, controlling narrative channels becomes increasingly important. Expect more AI companies to invest in owned media properties.
📈 Enterprise Model Competition Intensifies: Microsoft’s three new foundational models demonstrate that the enterprise AI battle is shifting from API access to in-house model development. Companies with existing enterprise relationships (Microsoft, Google) have distribution advantages that pure-play AI labs must overcome.
📉 AI Infrastructure Environmental Scrutiny: Meta’s natural gas-powered data center highlights the growing tension between AI expansion and environmental commitments. As AI power consumption becomes more visible, expect increased regulatory and public pressure on sustainable AI infrastructure.
📈 AI Supply Chain Security: The Mercor/LiteLLM incident follows the Anthropic GitHub takedown error, revealing systemic vulnerabilities in AI development infrastructure. Security practices have not kept pace with deployment velocity—a gap that will likely attract regulatory attention.
This briefing is generated automatically every 6 hours. For real-time updates, follow our AI news feed.