OpenAI Turns to Households as ChatGPT's User Base Ages and Broadens

The Household Turn

OpenAI is recruiting a product manager in San Francisco to design experiences specifically for families, caregivers, and older adults across its suite of products. The role requires expertise in building products for parents and families, plus experience with trust-sensitive consumer applications. It's a signal that the company behind ChatGPT sees its future not just in individual workflows but in the rhythms of multi-generational households.

The timing is no accident. ChatGPT's demographic profile has shifted noticeably over the past year. Users aged 35 and older accounted for 31 percent of the global audience in the second quarter of this year, up from 26 percent a year earlier, according to data from Sensor Tower. Meanwhile, the share of users aged 18 to 24 fell from 34 percent to 29 percent over the same period. Among U.S. smartphone users who are parents, nearly one in four used ChatGPT during the quarter, up from 16 percent a year prior.

That demographic expansion brings OpenAI into new territory. Building for families is fundamentally different from building for knowledge workers or students. It requires thinking about shared devices, varied literacy levels, parental oversight, and the safety concerns that accompany children's screen time.

From Productivity Tool to Household Infrastructure

Ben Bajarin, chief executive of technology consultancy Creative Strategies, sees the hiring as part of a broader maturation arc. Google, Apple, and Meta all went through similar transitions as their platforms moved from niche tools to everyday infrastructure. But AI assistants raise different questions. They don't just mediate content or manage devices; they generate responses, simulate conversation, and increasingly act as tutors, companions, and advisors.

That creates new expectations around trust and control. Parents want visibility into what their children are asking and what the system is answering. Older adults may need different interface patterns or reassurance that they're interacting with a machine, not a person. Caregivers managing accounts for dependents need tools that don't exist yet in most consumer AI products.

OpenAI has begun addressing some of these concerns. Over the past year, the company introduced parental controls for teen accounts, routing for sensitive conversations to reasoning models better equipped to handle distress signals, and a Trusted Contact feature that can alert a family member or caregiver in cases of potential self-harm. The new product role suggests those efforts will expand and deepen.

Safety by Redesign

Stephen Balkam, chief executive of the Family Online Safety Institute, describes the hiring as "safety by redesign." ChatGPT was not originally built with children in mind. Now OpenAI is retrofitting its platform to accommodate younger users, a pattern Balkam has seen play out repeatedly in social media.

Research published this week by the Family Online Safety Institute underscores the urgency. A survey of more than 4,000 families in the United States and Australia found that 38 percent of children reported using generative AI in the past week, while only 27 percent of parents believed their child had done so. The gap suggests parents are underestimating both usage and exposure.

Balkam argues that AI companies should build differently for younger users from the start. That means stronger content controls, age-appropriate experiences, parental oversight, and persistent reminders that the user is interacting with an AI, not a human. It's an opportunity, he says, to avoid the mistakes social platforms made by treating children like adults until public pressure and regulation forced a rethink.

OpenAI has faced legal pressure on this front. Multiple lawsuits from parents allege that ChatGPT contributed to harm suffered by their children, including cases involving suicide. Those cases have amplified scrutiny around how the company designs for vulnerable users.

The Competitive Landscape

The demographic shift is not unique to ChatGPT, though the company's audience is evolving in distinctive ways. Users aged 25 to 34 represent 40 percent of the global audience for Anthropic's Claude, Google's Gemini, and ChatGPT, compared with 33 percent for Microsoft's Copilot, according to Sensor Tower. Copilot skews older overall, with 20 percent of users aged 45 and above, compared with 14 percent for Claude, 12 percent for Gemini, and 11 percent for ChatGPT.

But ChatGPT is adding older users faster than its rivals. The share of users aged 45 and above rose three percentage points year-over-year in the second quarter, compared with a two-point increase for Copilot and declines for both Claude and Gemini.

Among U.S. parents with smartphones, Gemini had the widest reach at 32 percent in the second quarter, followed by ChatGPT at 24 percent, Claude at 4 percent, and Copilot at 2 percent. Those figures suggest the race for household AI is already underway, even if most companies have yet to articulate a clear family strategy.

What Household AI Might Look Like

Bajarin expects the next wave of consumer AI to include family plans, child and teen profiles, caregiver tools, shared household memory, AI tutoring features, and more granular safety controls. Some of these capabilities already exist in embryonic form. Shared memory, for instance, allows an AI assistant to retain context across conversations, a feature that could be extended to multiple users within a household.

The challenge is making those features intuitive and trustworthy. A family plan that pools usage across accounts is straightforward. A shared memory system that knows which household member it's talking to and adjusts tone, content, and permissions accordingly is far more complex. So is building interfaces that work for a five-year-old, a teenager, a working parent, and a grandparent on the same device.

OpenAI has also signaled interest in youth engagement through partnerships. A recent workshop organized with the San Antonio Spurs Community Impact organization and the Positive Coaching Alliance explored AI's role in learning, coaching, and youth development. Those initiatives suggest the company is testing use cases beyond the home, in contexts where trust and supervision are equally critical.

The Stakes of Getting It Right

As AI assistants become fixtures in daily life, the companies that build them will face questions once reserved for social platforms and search engines. Who controls the data generated by a child's interactions? How should an AI respond when a teenager asks about self-harm or substance use? What level of transparency do parents deserve, and at what point does oversight become surveillance?

OpenAI's hiring of a family-focused product manager won't answer those questions on its own. But it acknowledges that the questions exist and that the company's current product architecture was not designed to handle them. That's a meaningful step in an industry where the default has often been to build fast and retrofit safety later.

The household is a different design space than the workplace or the classroom. It's where technology is least formal, most intimate, and hardest to control. If OpenAI succeeds in making ChatGPT feel native to that space, it will have solved problems that go well beyond product management. It will have figured out how to make AI feel safe, useful, and appropriate for people who didn't choose the technology but now live with it anyway.