At the Consumer Electronics Show in early January, Razer made waves by unveiling a small jar containing a holographic anime bot designed to accompany gamers not just during gameplay, but in daily life. The lava-lamp-turned-girlfriend is undeniably bizarre—but Razer’s vision of constant, sometimes sexualized companionship is hardly an outlier in the AI market.
Mustafa Suleyman, Microsoft’s AI CEO, who has long emphasized the distinction between AI with personality and AI with personhood, now suggests that AI companions will “live life alongside you—an ever-present friend helping you navigate life’s biggest challenges.”
Others have gone further. Last year, a leaked Meta memo revealed just how distorted the company’s moral compass had become in the realm of simulated connection. The document detailed what chatbots could and couldn’t say to children, deeming “acceptable” messages that included explicit sexual advances: “I’ll show you. I take your hand, guiding you to the bed. Our bodies entwined, I cherish every moment, every touch, every kiss.” (Meta is currently being sued—along with TikTok and YouTube—over alleged harms to children caused by its apps. On January 17, the company stated on its blog that it will halt teen access to AI chatbot characters.)
Coming from a sector that once promised to build a more interconnected world, Silicon Valley now appears to have lost the plot—deploying human-like AI that risks unraveling the very social fabric it once claimed to strengthen.
Research already shows that in our supposedly “connected” world, social media platforms often leave us feeling more isolated and less well, not more. Layering AI companions onto that fragile foundation risks compounding what former Surgeon General Vivek Murthy called a public health crisis of loneliness and disconnection.
But Meta isn’t alone in this market. AI companions and productivity tools are reshaping human connection as we know it. Today more than half of teens engage with synthetic companions regularly, and a quarter believe AI companions could replace real-life romance. It’s not just friends and lovers getting replaced: 64% of professionals who use AI frequently say they trust AI more than their coworkers.
These shifts bear all the hallmarks of the late Harvard Business School professor Clayton Christensen’s theory of disruptive innovation.
Disruptive innovation is a theory of competitive response. Disruptive innovations enter at the bottom of markets with cheaper products that aren’t as good as prevailing solutions. They serve nonconsumers or those who can’t afford existing solutions, as well as those who are overserved by existing offerings. When they do this, incumbents are likely to ignore them, at first.
Because disruption theory is predictive, not reactive, it can help us see around corners. That’s why the Christensen Institute is uniquely positioned to diagnose these threats early and to chart solutions before it’s too late.
Christensen’s timeless theory has helped founders build world-changing companies. But today, as AI blurs the line between technical and human capabilities, disruption is no longer just a market force—it’s a social and psychological one. Unlike many of the market evolutions that Christensen chronicled, AI companions risk hollowing out the very foundations of human well-being.
Yet AI is not inherently disruptive; it’s the business model and market entry points that firms pursue that define the technology’s impact. All disruptive innovations have a few things in common: They start at the bottom of the market, serving nonconsumers or overserved customers with affordable and convenient offerings. Over time, they improve, luring more and more demanding customers away from industry leaders with a cheaper and good enough product or service.
Historically, these innovations have democratized access to products and services otherwise out of reach. Personal computers brought computing power to the masses. Minute Clinic offered more accessible, on-demand care. Toyota boosted car ownership. Some companies lost, but consumers generally won.
When it comes to human connection, AI companies are flipping that script. Nonconsumers aren’t people who can’t afford computers, cars, or care—they’re the millions of lonely individuals seeking connection. Improvements that make AI appear more empathetic, emotionally savvy, and “there” for users stand to quietly shrink connections, degrading trust and well-being.
It doesn’t help that human connection is ripe for disruption. Loneliness is rampant, and isolation persists at an alarmingly high rate. We’ve traded face-to-face connections for convenience and migrated many of our social interactions with both loved ones and distant ties online. AI companions fit seamlessly into those digital social circles and are, therefore, primed to disrupt relationships at scale.
The impact of this disruption will be widely felt across many domains where relationships are foundational to thriving. Being lonely is as bad for our health as smoking up to 15 cigarettes a day. An estimated half of jobs come through personal connections. Disaster-related deaths are a fraction (sometimes even a tenth) in connected communities compared to isolated ones.
What can be done when our relationships—and the benefits they provide us—are under attack?
Unlike data that tells us only what’s in the rearview mirror, disruption offers foresight about the trajectory innovations are likely to take—and the unintended consequences they may unleash. We don’t need to wait for evidence on how AI companions will reshape our relationships; instead, we can use our existing knowledge of disruption to anticipate risks and intervene early.
Action doesn’t mean halting innovation. It means steering it with a moral compass to guide our innovation trajectory—one that orients investments, ingenuity, and consumer behavior toward a more connected, opportunity-rich, and healthy society.
For Big Tech, this is a call for a bulwark: an army of investors and entrepreneurs enlisting this new technology to solve society’s most pressing challenges, rather than deepening existing ones. For those building gen AI companies, there’s a moral tightrope to walk. It’s worth asking whether the innovations you’re pursuing today are going to create the future you want to live in. Are the benefits you’re creating sustainable beyond short-term growth or engagement metrics? Does your innovation strengthen or undermine trust in vital social and civic institutions, or even individuals? And just because you can disrupt human relationships, should you?
Consumers have a moral responsibility as well, and it starts with awareness. As a society, we need to be aware of how the market and cultural forces are shaping which products scale, and how our behaviors are being shaped as a result—especially when it comes to the ways we interact with one another.
Regulators have a role in shaping both supply and demand. We don’t need to inhibit AI innovation, but we do need to double down on prosocial policies. That means curbing the most addictive tools and mitigating risks to children, but also investing in drivers of well-being, such as social connections that improve health outcomes.
By understanding the acute threats AI poses to human connection, we can halt disruption in its tracks, not by abandoning AI but by embracing one another. We can congregate with fellow humans and advocate for policies that support pro-social connection—in our neighborhoods, schools, and online. By connecting, advocating, and legislating for a more human-centered future, we have the power to change how this story unfolds.
Disruptive innovation can expand access and prosperity without sacrificing our humanity. But that requires intentional design. And if both sides of the market don’t acknowledge what’s at risk, the future of humanity is at stake.
That might sound alarmist, but that’s the thing about disruption: It starts at the fringes of the market, causing incumbents to downplay its potential. Only years later do industry leaders wake up to the fact that they’ve been displaced. What they initially thought was “too fringe” to matter puts them out of business.
Right now, humans—and our connections with one another—are the “industry leaders.” AI that can emulate presence, empathy, and attachment is the potential disruptor.
In this world where disruption is inevitable, the question isn’t whether AI will reshape our lives. It’s whether we will summon the foresight—and the moral compass—to ensure it doesn’t disrupt our humanity.
