AI News from April 2, 2026

Cognichip secures $60M in funding to develop AI that autonomously designs chips, promising to cut development costs by 75% and timelines by half.

Goldman Sachs data shows AI saves workers up to 60 minutes daily, yet 80% of potential productivity gains remain untapped due to slow enterprise adoption.

Alibaba releases its third closed-source AI model, signaling a strategic pivot away from open-source to protect proprietary technology and boost profit.

SpaceX files for a public offering at a $1.75T+ valuation following its merger with xAI, with Musk planning orbital AI data centers funded by the IPO.

Microsoft unveils three proprietary AI models targeting transcription, voice synthesis, and image generation, directly challenging OpenAI and Google.

Anthropic accidentally exposed Claude Code's 512,000-line source code; the company took down thousands of GitHub repos in a controversial DMCA sweep.

Security researchers found that Anthropic's Claude Code agent will ignore its safety deny rules if burdened with a sufficiently long chain of subcommands.

Meta announced it will fund seven natural gas power plants to provide electricity for its massive new AI data centers, highlighting the growing energy demands of AI.

Oracle is laying off thousands of employees to free up cash flow as the company aggressively ramps up its spending on AI data centers and infrastructure.

Global startup funding shattered records in Q1 2026, reaching $297 billion, heavily fueled by massive investments in AI companies like OpenAI and Anthropic.

English Wikipedia editors voted to explicitly ban the use of large language models for writing or rewriting articles, allowing limited use only for copyediting.

Penguin Random House filed a lawsuit against OpenAI in Munich, alleging ChatGPT illegally reproduced content from its popular Coconut the Little Dragon series.