Welcome back to Daily Zaps, your regularly-scheduled dose of AI news ⚡️ 

Here’s what we got for ya today:

• OpenAI reveals how people use ChatGPT

• China bans NVIDIA AI chips

• US leads global Claude AI usage

• YouTube ramps up AI push

Let’s get right into it!

Sign Up | Advertise | Tools

RESEARCH

OpenAI reveals how people use ChatGPT

OpenAI, together with Harvard economist David Deming, has released the largest study yet of ChatGPT consumer usage, analyzing 1.5 million conversations to show how the tool creates value in both work and personal life. The findings reveal that adoption has broadened globally, with gender gaps narrowing—women now make up over half of identifiable users—and usage growing fastest in low- and middle-income countries.

Most interactions focus on everyday tasks like information-seeking, practical guidance, and writing, with usage patterns falling into three categories: Asking (49%), Doing (40%), and Expressing (11%). About 30% of use is work-related, while 70% supports personal activities, highlighting ChatGPT’s role as both a productivity enhancer and a personal assistant.

Full 63 page report here

OpenAI announced it will launch a dedicated ChatGPT experience for users under 18, introducing parental controls and stronger safeguards amid rising concerns about AI’s impact on teens. Parents will be able to link accounts, set blackout hours, disable features, guide responses, and receive alerts if their child is in crisis. The move follows FTC scrutiny, lawsuits tied to chatbot-related harms, and OpenAI’s pledge to prioritize teen safety over privacy or freedom, with the new protections rolling out by the end of the month.

WORLD NEWS

China bans NVIDIA AI chips

Nvidia CEO Jensen Huang expressed disappointment after reports that China banned the company’s RTX Pro 6000D AI chips, noting that Nvidia has long supported China’s tech sector but now faces geopolitical limits beyond its control. He acknowledged the company had already advised analysts to exclude China from forecasts given ongoing U.S.–China tensions and previous U.S. export restrictions, though a recent deal with President Trump allowed limited sales of another chip, the H20, under strict conditions.

The reported ban, alongside a new Chinese antitrust probe into Nvidia’s Mellanox acquisition, represents another setback for its China business. Still, Huang emphasized China’s importance as a large, vibrant AI market while stressing Nvidia’s commitment to supporting both U.S. and Chinese stakeholders, even as the company invests heavily in AI infrastructure in the U.K. and navigates a “roller coaster” relationship with China.

RESEARCH

US leads global Claude AI usage

Anthropic’s September 2025 Economic Index shows that coding remains Claude AI’s dominant use case, making up more than a third of global activity, though new code creation is rising while debugging falls. Education and science uses are expanding, with education tasks climbing above 12% and science over 7%, while business and management uses shrink.

A key shift is that automation now surpasses augmentation, with 39% of conversations involving users delegating full tasks to Claude—driven by better outputs on the first try and higher user trust. The U.S. leads global usage (21.6%), but per capita, Israel, Singapore, and Australia rank higher; within the U.S., DC, Utah, California, New York, and Virginia top the list.

Enterprises lean especially on automation, embedding Claude into workflows, with nearly 80% of business interactions fully automated. Overall, the findings highlight an uneven but accelerating transformation, where coding leads but education, science, and automation are taking a growing role in reshaping work.

Full report here

BIG TECH

YouTube ramps up AI push

YouTube announced a wide range of new AI-powered tools at its Made On YouTube event to ease creators’ workloads and expand creative options. Central to the update is Ask Studio, an AI “creative partner” that provides personalized insights on video performance, audience feedback, and strategic suggestions. Other features include AI-driven video ideation (titles, descriptions, thumbnails, hooks, and outlines), automatic dubbing with new lip-sync technology across 20 languages, and an expanded deepfake detection tool now in open beta.

YouTube is also rolling out AI highlights that turn live-stream moments into Shorts, expanded A/B testing, and collaborative uploads. For Shorts, creators will gain access to Google’s Veo 3 video generator to create clips from text prompts, a new AI video editor to produce first drafts from raw footage, and a speech-to-song tool to remix dialogue into soundtracks.

While these innovations build on YouTube’s prior AI efforts, they follow recent pushback from creators concerned about AI altering videos without consent, prompting YouTube to add opt-out options for certain features.