As the first month of 2026 wraps up, I’ve been thinking about early signals and what the rest of this year will look like as AI adoption and innovation accelerate.

We’re entering this next phase at a time when there are serious social, political, and economic challenges. Trust is fragile and technology increasingly shapes how people think, decide and relate to one another. The bonds that hold humanity together are strained, and it’s more important than ever that AI is designed to be human-centered, to support people, not overwhelm them.

The past year has been dominated by speed and scale. The race has been about getting AI to mainstream audiences as quickly as possible, not about how to improve people’s daily lives.

2026 will reflect a necessary shift in focus.

As AI becomes embedded in daily life, the next chapter won’t be defined by smarter, faster models alone, but by how meaningfully AI improves people’s lives. That means being human-centered, understanding real context, prioritizing individual needs, and being honest about what AI can and can’t do. As leaders, our responsibility isn’t just to drive adoption, but to educate and build trust along the way.

Human-centered AI is less about what the system can do and more about how it fits into human lives. It should make people more capable, not sidelined. Human-centered AI recognizes that cognition and emotion are intertwined. It is designed to support autonomy, consent, and boundaries especially around memory, personalization, and data. It understands situations, not just inputs.

Here are a few trends I believe will accelerate in the year ahead:

IQ with EQ: Building trust in AI requires a balance of emotional intelligence and technical intelligence. While much of the industry is focused on raw reasoning and problem-solving, emotional intelligence is what turns AI from transactional to genuinely relational. AI can infer and perceive what you need as it learns from you, so you don’t need to have the perfect prompt. Better context leads to better results. That belief is core to how we built Pi, as an engaging, empathetic, conversational, and contextually aware AI that empowers people rather than overwhelms them. That balance is tough to achieve. We’re aiming for emotional intelligence without sycophancy.

AI literacy will drive mass adoption: Widespread adoption won’t come from incentives alone, it will come from literacy. People need clear, accessible ways to understand how AI works, what it’s good at, and where its limits are. AI literacy is growing, but it needs to grow faster and be easier for consumers. Advancing tech literacy is something that I’m passionate about and have driven forward since my work at Mozilla. We will do our part to advance AI literacy as a Public Benefit Corporation, and support efforts like National AI Literacy Day.

The proliferation of AI-powered devices: This year, we’ll see a wave of new AI-enabled devices that change not just how we use systems, but how we engage with them socially. We’ve already seen indications of the proliferation of AI-powered devices and wearables from companies like Apple, Meta, and OpenAI. After leading the wearables team at Nokia and spending years working on the development of devices in my career, it’s exciting to see AI finally make interfaces people talked about a decade ago genuinely viable today.

Voice-first interaction will grow: AI is moving from a text-first paradigm to voice-first experiences, bringing human to computer interactions to a whole new level. Text and typing will continue to be important for accessibility, but voice interactions unlock entirely new ways to elevate the human experience and amplify our capabilities beyond what we can imagine today. Voice-powered experiences will help AI give people superpowers. As my co-founder, Reid Hoffman puts it, becoming “voicepilled” is the realization that using your voice to interact with technology unlocks new ways to amplify your ability.” We see strong engagement and appreciation for the voice features in Pi today and we’re experimenting with new innovations that we will ship this year.

More personalization: AI will become increasingly personalized, which ultimately means deeper context. Again, better context leads to better results. This personalization will require more of people’s data and raises an important question: who do people trust with their data, and why? As personalization increases, transparency, alignment of incentives, and clarity about whose interests AI serves will matter more than ever.

Trustworthy agentic frameworks: As AI systems move from answering questions to taking actions, what people talk about as agentic AI, strong frameworks and guardrails become essential. At Inflection AI, we start by applying a simple rubric about when and how humans should remain in the loop, especially when actions are irreversible or have significant impact. This is an area where the leaders in the industry need to work together on innovation, restraint and thoughtful design as things evolve.

We’re still early in AI adoption at the beginning of 2026. I’m excited about the positive ways AI will augment people and help unlock capabilities we’re only beginning to imagine.

What trends do you think we’ll see in AI this year?