At CES 2026, weird AI gadgets steal the show—from holographic helpers to robotic pets

AI is everywhere at CES—especially in the strangest products

CES 2026 in Las Vegas delivered the expected parade of upgrades from major brands, but the most attention-grabbing announcements often came from products that felt equal parts playful, unsettling, and oddly plausible. As artificial intelligence becomes a default marketing ingredient, companies are experimenting with form factors that would have sounded like jokes a few years ago: holographic assistants, emotional-support robots, and countertop appliances that claim machine-learning benefits.

One headline concept: a small holographic desk companion from Razer’s Project AVA line, presented as a character-based assistant that can help with gaming, productivity, organization, and even personal advice. The pitch leans heavily on “presence”—expressive movement, lip-syncing, and a sense of being watched. That last point matters: the device concept includes camera-based tracking, which raises immediate questions about privacy, local processing, and who has access to the data.

From elder-care robotics to AI noise reduction

Another oddball crowd-pleaser was an AI “pet” panda robot positioned as companionship tech for older adults. The idea is simple: people bond with responsive, tactile feedback. By adding sensors and memory of interactions, the robot aims to feel more personalized over time, potentially supporting well-being for users who are isolated. As with many “emotional AI” products, the challenge is proving meaningful benefits while avoiding manipulative design that encourages dependency.

Even kitchen appliances joined the AI boom. A countertop nugget ice maker was promoted with an AI feature intended to reduce noise by detecting freeze-ups and triggering preventative defrosting. It’s a small claim, but it points to a broader CES theme: AI is migrating from big promises about general intelligence into narrow, practical tasks—smoother mechanics, better energy use, and predictive maintenance.

What matters now: usefulness, software maturity, and trust

CES has always been a mix of serious roadmaps and experimental theater, and 2026’s show leaned hard into the experiment. But as AI becomes routine, the real differentiator is less about whether a product has AI and more about whether it uses it responsibly. Observers increasingly focus on software maturity—how reliably features work in real homes—and on trust issues such as on-device processing, transparent data policies, and safe defaults.

In other words, AI’s novelty is fading, but scrutiny is rising. The most memorable CES demos may be the weirdest, yet the products likely to endure are those that use AI quietly, measurably, and with minimal risk—improving daily life without demanding constant surveillance or rewriting the boundaries of consent.