OpenAI and Anthropic Burned $27 Million Fighting One Lawmaker and Came Out Even
1. OpenAI and Anthropic Spent $27 Million Against One New York Lawmaker. It Ended in a Draw. Alex Bores, a New York state Assemblyman, narrowly lost the Democratic primary for New York's 12th Congressional district.
2. Computer use lands in Google's cheap Flash tier, as two new benchmarks map where these agents break Google moved computer use into Gemini 3.5 Flash, the model it sells on speed and price. The capability had lived in a standalone Gemini 2.5 computer use model.
3. A24's fans spent years championing the anti-Hollywood studio. Its new backer is Google DeepMind. The fans who turned A24 into a brand did it by treating the studio as the opposite of everything they distrusted about big Hollywood.
In Brief
- US memory-chip maker's quarterly profit jumps to $28.2 billion A US memory supplier rode the AI-driven memory shortage to $41.45 billion in revenue, up from roughly $10 billion a year earlier. Profit climbed from $1.88 billion to $28.2 billion year-over-year.
- Cerebras stock falls after first earnings as public company Cerebras forecast a narrower gross margin in its core business, sending shares down in its debut report since the IPO. The CEO said investors misread the margin guidance.
- Google AI researchers Jonas Adler and Alexander Pritzel leave for Anthropic Two senior Google AI researchers departed for Anthropic, extending an exodus that already took Noam Shazeer and AlphaFold's John Jumper. The departures hit Google's core research bench.
- OpenAI launches program to find and patch open-source bugs OpenAI started an initiative that uses its models to detect and fix vulnerabilities in open-source projects, then submit patches upstream. The effort targets widely used dependencies.
- Engineers make up a growing share of new hires despite AI layoff narrative SignalFire data shows engineering roles are among the most resilient to AI displacement, gaining share of new hires rather than shrinking. The finding contradicts predictions that AI would cut engineering headcount first.
- Facebook tests a standalone AI companion app for creators Facebook began testing a separate app built around its creator assistant, letting select creators interact with an AI companion. The assistant launched recently inside the main app.
- India's MoEngage buys technology to assign one AI agent per customer MoEngage made an all-cash acquisition for technology that pairs individual customers with dedicated marketing AI agents. The company is betting marketing automation will run on millions of per-user agents.
- Google Search now stores your media uploads to train AI; here's how to opt out Google's updated Search history retains media from interactions, such as images submitted to reverse image search, and uses them to train its models. Users can disable the retention in settings.
- Qwen releases language world models for simulating agent environments Alibaba's Qwen team introduced Qwen-AgentWorld, including 35B and 397B mixture-of-experts models that predict environment dynamics from observations and actions. The models aim to let agents reason and plan against a learned simulator instead of live environments.
Don't miss what's next. Subscribe to AI News Digest: