Daily AI News | ChatGPT ad testing expands to the UK, Japan, and three more markets
# Daily AI News | ChatGPT ad testing expands to the UK, Japan, and three more markets
## Lead Story
### ChatGPT ad testing expands to the UK, Japan, and three more markets
OpenAI expanded its ChatGPT ad test to the United Kingdom, Mexico, Brazil, Japan, and South Korea. The company says ads will be clearly labeled, will not influence answers, and will retain privacy protections and user controls.
**AI take:** The move takes ads into more major markets, accelerating monetization of free access while putting answer independence and privacy promises under a broader real-world test.
## More News
### OpenAI brings Daybreak cyber models to Amazon Bedrock
OpenAI and AWS are making Daybreak Red and Daybreak Blue available through Amazon Bedrock to eligible, enrolled customers for enterprise cybersecurity workflows. Access remains subject to qualification and governance controls.
**AI take:** Frontier cyber models now have a clearer enterprise adoption path through a major cloud, while their strongest capabilities remain tied to controlled access.
### NVIDIA releases Nemotron 3.5 Lightning and open-source Switchyard
NVIDIA released Nemotron 3.5 Lightning for long-running agent tasks and introduced the open-source NeMo Switchyard routing library, which can direct agent workloads across different models. Model weights are available for download.
**AI take:** Agent competition is expanding from individual models to execution and routing stacks, where assigning each task to the right model can become a product advantage.
### OpenAI launches a ChatGPT desktop app for Linux
TechCrunch reported that OpenAI launched a dedicated ChatGPT desktop app for Linux, filling the remaining major desktop-platform gap and supporting ChatGPT and coding-oriented workflows.
**AI take:** A native Linux client reduces workflow friction and strengthens Codex access.
### Anthropic adds machine-readable watermarks to Claude output
Anthropic said it will add machine-readable watermarks to Claude-generated content and extend watermarking support to older models, helping platforms and users identify AI-generated material.
**AI take:** Adoption and resilience to editing will determine whether the watermark works.
### Two-month-old River AI raises $1.1 billion
TechCrunch reported that River AI, founded by xAI co-founder Igor Babuschkin, raised $1.1 billion in a round led by General Catalyst. The roughly two-month-old company is focused on personal agents.
**AI take:** The capital is striking, but durable products still have to justify it.
## Rumors
