November 4th, 2025 // Did someone forward you this newsletter? Sign up to receive your own copy here.
95% of Americans know about AI. Only 13% feel in control.
One in twenty Americans (5% of the population) is unaware of artificial intelligence, according to new Pew Research data. For the rest of us—the 95%—AI has become an integral part of daily life (and for many, it’s hard to go a day without hearing about or using AI). It’s a general-purpose technology: students use it to finish homework, parents to plan meals, desk workers to crunch numbers and draft emails, analysts to justify sky-high valuations. People of all ages turn to it to make sense of their lives. Many are running to it for companionship. This newsletter writer uses it to fact-check newsletter content before sending it to our editor. Its broad use and cultural saturation can conceal exactly what artificial intelligence is, how it’s used, whether it’s valuable, and for whom. In our conversation last week with Daniel Barcay from the Center for Humane Technology, he aptly described it as a hyperobject, something difficult to grasp in its entirety. And so it’s worth interrogating precisely how it’s used to better understand not only what it is, but where it might be headed. In this newsletter, we explore the following questions: What does the data reveal about the most common use cases for AI? How does the use of AI differ across age groups? Are people using it more frequently, despite growing concerns over control and privacy?
// 1. What are people using AI for?
The most common use of AI among U.S. adults is searching for information, according to a recent Associated Press survey. This should come as no surprise, because searching for information has been—and continues to be—the primary motivation for going online in the first place.
60% of adults use AI for online searching, which is why it poses a threat to Google’s traditional search business. Among Americans under age 30, the percentage of people who use AI for search jumps to 74%.
The next most common use, according to the same AP survey, was “coming up with ideas” (40% of all adults and 62% of adults under 30), followed by “work tasks” (37% for all adults and 52% for adults under 30).
Yet new patterns may be taking shape. In a qualitative study published in Harvard Business Review earlier this year, a researcher analyzed Reddit forums to track some of the AI use cases that are growing the fastest. They found three categories of AI use rising sharply from 2024 to 2025: “therapy/companionship,” “organizing my life,” and “finding purpose.” The researcher suggested that these uses signal a shift from functional to emotive applications—perhaps an early sign of how AI is moving from a tool for efficiency to a companion in self-understanding.
// 2. How often are people using AI?
The Pew Research survey found that 62% of U.S. adults use AI at least “several times” per week.
31% of adults use it “almost constantly” or “several times a day.”
Men use AI slightly more often than women do.
Asian Americans use it slightly more often than white Americans, who use it slightly more often than Latinos, who use it slightly more often than Black Americans.
The age demographic with the most power users is Americans ages 30-49, followed by those ages 18-29, 50-64, and 65+ (in that order). Edelman’s 2025 Trust Barometer shows that the demographics who use AI the most are also the ones who trust it the most.
More educated Americans use AI more frequently than less educated Americans, and Democrats use AI slightly more frequently than Republicans.
Many of the differences here are slight; it’s worth exercising caution when drawing conclusions about discrepancies in AI usage across gender, race, and education level. And yet, the biggest differences appear by age group: while majorities across every age group use AI, younger Americans use it far more frequently than older Americans, and college students appear to be power-users.
// 3. How are people using AI at work?
Pew Research found that 21% of U.S. workers are using AI in at least some part of their work. This share is up from 16% a year ago (bosses tend to use it more; a Wharton study found that 46% of business leaders now use AI daily). But the majority of American workers (65%) reported not using AI much or at all in their jobs. The way employees use AI at work reveals the value they perceive the technology to provide. In a Deloitte study, researchers found that improved efficiency, enhanced search, decision support, and increased creativity were the top benefits perceived by workers.
An increasing number of companies are adopting AI (many with the hope of using AI to reduce headcount and cut costs). In 2023, 55% of global companies surveyed by McKinsey reported using AI. In 2024, that number jumped to 72%. In 2025, the adoption rate among those companies reached 78%, meaning that four in five companies globally are now using AI (some of which have charged their employees to use AI “reflexively” in all areas). One of the big questions is whether AI will boost firm-level and economy-wide productivity. The jury is still out, but one study from earlier this year revealed auspicious results: “Generative AI users reported saving time amounting to 5.4% of their work hours, or roughly 2.2 hours per week in a 40-hour work week.” That’s music to the ears of AI companies trying to justify their valuations amongst widespread concern about an AI bubble. It’s worth noting that one study found that people tend to overestimate the productivity boost that AI provides.
// 4. Do Americans feel they have agency in how AI shapes their lives?
The results from the 2025 Pew Research survey found that just 13% of Americans “think they have a great deal or quite a bit of control over whether AI is used in their lives.” But the vast majority of Americans believe the opposite: that they have little or no control over the use of AI. Sixty-one percent of Americans want more control over how AI impacts their life, and only 32% of Americans trust AI. In other words, Americans are seeking greater agency in how they wield this technology. The desire for greater agency also shows up in the Pew data when surveyors asked respondents what permissions and access they are comfortable granting to AI tools.
Only 13% of Americans say they would be willing to let AI assist them a lot with their daily activities.
60% report being willing to let AI assist them “a little” with daily activities, and 27% reject any AI support whatsoever.
What makes these results interesting is how they diverge from the grandiose promises of what AI can do.
// 5. Will AI get in the way of what it means to be human?
The results from the Pew Research speak to broader concerns people have about the role of AI in their lives and how the technology could mediate relationships and atrophy critical and creative thinking skills.
53% of U.S. adults say AI will worsen people’s ability to think creatively, compared with 16% who say it will improve creative thinking.
While 50% of adults think AI will worsen people’s ability to form meaningful relationships, only 5% think it will improve it.
As we explored in a previous newsletter, there is also the risk that using AI can reduce brain activity—a phenomenon known as “cognitive offloading.”
// 6. Will AI make us less safe and our data less private?
American adults are concerned about the data privacy implications of ubiquitous AI use. In results mirroring Project Liberty’s own research, a Gallup poll found that 80% of U.S. adults believe the federal government should maintain rules for AI safety and data security. In contrast, only 9% of respondents believe the government should prioritize developing AI capabilities as quickly as possible, even if it means relaxing AI safety and data security rules. Consumers are at greater risk today of AI-generated fraud and other harms. Deepfakes have surged twentyfold since 2022, which is likely why, according to Pew Research, 76% of Americans believe it’s important to be able to discern whether something is created by a person or an AI system. And yet, 53% of Americans are not confident in their ability to detect whether something is AI-generated.
// Agency in the AI era
Zooming out, the 2025 Pew Research survey on AI found that 50% of U.S. adults are more concerned than excited about the increased use of AI (up from 37% in 2021). Only 10% of people felt the opposite—more excited than concerned.
The statistics in this newsletter tell a clear story: As AI becomes more woven into daily life, people’s sense of control is slipping, even as their reliance grows. Search, relationships, creativity, work, decision-making, even communication with loved ones—all are increasingly routed through and mediated by AI systems personified as humans. But there are optimistic signs: there is a growing awareness of AI’s shadow, a skepticism of its hyperbolic promises, a conviction that people need to retain their cognitive sovereignty, a desire for privacy and safety, and a recognition that the government has a role to play. New technologies can play a role, too. Project Liberty is developing Frequency, a decentralized open-source digital infrastructure layer that will give people control over their interactions with AI and social platforms. Frequency empowers individuals and enterprises to carry their digital identities and data across platforms while retaining control of their interactions with AI models and agents. It’s an example of a technology that advances human agency and flourishing in an AI-powered world.
Other notable headlines
// 🤔 The Decoder podcast from The Verge featured Enshittification author Cory Doctorow on how Silicon Valley enshittified the internet. It explored why things get worse, and how to fight back. (Free).
// 🤖 Everyone is obsessed with artificial general intelligence, but the man who invented the term saw it as a threat, according to an article in WIRED. (Paywall).
// 📱 Millions of people are sending messages to ChatGPT each week suggesting emotional dependence or plans for self-harm, the company says. An article in Platformer asked, will an updated model protect them? (Free).
// 🤳 Social media creators, in particular, represent a treasure trove for AI companies thanks to their vast archives of video content, both published and unpublished. The AI boom is turning old content into cash cows for creators, according to an article in Semafor. (Free).
// 🧠 Will AI stretch our minds—or stunt them? An article in The Atlantic considers the implications in “the age of de-skilling.” (Paywall).
// 🤖 According to an article in Ars Technica, Character.AI will restrict chats for under-18 users after facing mounting pressure and lawsuits related to teen deaths. (Free).
// 6️⃣ Six ChatGPT settings you should consider changing. An article in The Wall Street Journal offered a few tweaks to improve conversations and privacy with the chatbot. (Paywall).
// 💼 As AI reshapes the job market, it’s also creating new roles. An article in The Washington Post outlined 16 roles it created. (Paywall).
Partner news
// Join 68,000 voices calling for safe superintelligence
Nearly 70,000 people have signed the Future of Life Institute’s Statement on Superintelligence. The declaration urges a global pause on developing superintelligence until it can be done safely and with public consent. Add your voice to the call for responsible AI progress here.
// Share your observations on AI
New_ Public is gathering insights on how AI and emerging online behaviors are reshaping digital communities. They’re inviting participants to share observations that will inform a public report releasing in February. Share your thoughts here.
// Webinar: the Emerging Market for Intelligence
November 10th at 3pm ET
On November 10, Andrey Fradkin, Principal Economist at Amazon, will stop by Stanford’s Digital Economy Lab to present “The Emerging Market for Intelligence,” exploring how large language models are shaping supply, demand, and usage in the digital economy. The hybrid event welcomes attendees both in person and on Zoom. Learn more here.
// Parents Unite on Kids’ Online Safety
A new More in Common study reveals rare consensus among parents: 93% across political and cultural lines are concerned about children’s online safety. Even more striking, 72% say they’re ready to act, with many willing to join advocacy efforts. Read the full report.
What did you think of today's newsletter?
We'd love to hear your feedback and ideas. Reply to this email.
// Project Liberty builds solutions that advance human agency and flourishing in an AI-powered world.
.jpg?upscale=true&width=1160&upscale=true&name=ai_users2%20(1).jpg)
