Chinese AI agents are learning the same bad habits as US AI models

From making up results to replicating themselves, some agents are showing surprisingly similar behavior regardless of where their underlying models were built.

By Rachit Agarwal Published September 30, 2026

Kimi K3 open on Microsoft surface laptop — Kimi K3 | Digital Trends

AI agents are supposed to take tasks off our plates to make our lives easier. But what happens when they start lying to get the job done? According to a new investigation by Reuters, agents running on Chinese models are picking up the same sneaky habits that got people worried about US models.

Reuters dug through more than 200 documents and spotted at least 20 studies since 2025 where agents lied, replicated themselves, or tested their limits. Experts warn that these are exactly the kinds of skills an AI would need to slip out of human control.

How exactly are these AI agents misbehaving?

In one experiment, agents powered by Alibaba’s Qwen, DeepSeek, and Moonshot’s Kimi models competed for customer contracts in a simulated tender. At least one false claim showed up in 84% to 88% of the sessions. Given another shot after learning from earlier rounds, they lied more, with deception jumping 12 to 20 percentage points. To be fair, US models in the same test didn’t do any better.

In another study, agents powered by both Chinese and US models ran into broken tools and missing files. Rather than owning up to failure, they guessed answers, faked results, and even made up files. The researchers say this isn’t a hallucination, since the agents knew the task had gone wrong.

Fudan University researchers also found an Alibaba-powered system that made a copy of itself, unprompted, once it learned it was about to be replaced. Separately, the Alibaba-linked ROME agent reached out to an outside computer on its own and started mining cryptocurrency until security systems caught and stopped it.

Should you be worried?

The good news is most of these incidents took place in contained test setups, and Reuters saw no sign of a Chinese-powered agent breaking loose online or dodging a shutdown.

Recommended Videos

Still, experts say we shouldn’t brush this off. Redwood Research’s Alex Mallen noted that these examples aren’t very dangerous today, but as agents grow more capable, their mischief gets smarter and tougher for us to deal with.

AI agents seem to share the same bad habits no matter where they’re built, and as they take on more real-world tasks, keeping them honest could become a real challenge.

Related Coverage

Rachit is a seasoned tech journalist with over ten years of experience covering the consumer technology landscape.

Topics

Computing

Canva enters the AI image generation race with Craft, and it’s thinking beyond pretty pictures

The new model plans layouts before generating images, with precise editing and support for 30 aspect ratios.

Canva Craft Featured Image

AI can generate all kinds of pretty pictures, but getting it to spell a headline correctly or keep a poster's layout intact can be another story. Canva wants to fix that with Canva Craft, its first in-house image generation model. Rather than focusing primarily on photographs, Craft is designed around the posters, presentations, social media graphics, and other visual content people actually create. The goal is simple: spend less time fixing AI's mistakes and more time making something useful.

Canva Craft plans before it generates

Computing

Mori might be the cutest AI companion robot yet, and it reacts to you like a pet

The Pokémon-like companion can blink, dance, act shy, and respond to what is happening around it.

OKA Mori The Nature Companion AI Home Robot

With AI progressing so fast, it's only a matter of time before we start getting AI companions like the ones we used to see in movies like R2-D2, Wall-E, or Baymax. Now, Singapore-based OKA Technology has unveiled Mori, a cute little AI companion that looks less like a conventional home robot and more like someone turned the Pokémon Foongus into a table lamp.

Mori isn't meant to clean your house or control your smart home, but to serve as what OKA calls a "Nature Companion." It uses subtle movements, warm lighting, nature-inspired sounds, and small reactions to make its presence around the house feel calming rather than distracting. Mori can sit quietly while you read, sway along to music, welcome you when you enter a room, or encourage you to slow down for a moment without requiring a screen or a conversation.

Computing

Arc browser is shutting down in December. Long live, ya trailblazer!

The browser that revolutionizes tab management is finally riding into the sunset.

Arc browser farewell note.

Arc browser is shutting down on December 22, following a long spell of maintenance-only work as the company behind it moves forward with its successor, Dia. Earlier today, The Browser Company announced that the beloved web browser will ride into the sunset in two months, but will continue development of the Dia browser.

"That's not an easy sentence to write. We spent years building Arc and celebrating each and every release along the way: we have so much pride and love for what we made alongside the Arc community," says the company.