Are we ready to hand AI agents the keys?

Agents are already everywhere—and have been for many decades. Your thermostat is an agent: It automatically turns the heater on or off to keep your house at a specific temperature. So are antivirus software and Roombas. They’re all built to carry out specific tasks by following prescribed rules. But in recent months, a new class of agents has arrived on the scene: ones built using large language models. 

Tech CEOs and scholars alike are taking AI agents seriously. In this edition of What’s Next in Tech, examine whether we’re ready to allow this new type of agent into our everyday lives.

📩 We want to hear from you! Do you have a question about something you read, or a topic that we cover? Let us know at newsroom@technologyreview.com and we'll answer some of your questions in our next magazine issue.

We’re starting to give AI agents real autonomy, and we’re not prepared for what could happen next.

Operator, an agent from OpenAI, can autonomously navigate a browser to order groceries or make dinner reservations. Systems like Claude Code and Cursor’s Chat feature can modify entire code bases with a single command. Manus, a viral agent from the Chinese startup Butterfly Effect, can build and deploy websites with little human supervision. Any action that can be captured by text—from playing a video game using written commands to running a social media account—is potentially within the purview of this type of system.

LLM agents don’t have much of a track record yet, but to hear CEOs tell it, they will transform the economy—and soon.

Scholars, too, are taking agents seriously. “Agents are the next frontier,” says Dawn Song, a professor of electrical engineering and computer science at the University of California, Berkeley. But, she says, “in order for us to really benefit from AI, to actually [use it to] solve complex problems, we need to figure out how to make them work safely and securely.”

That’s a tall order. Because like chatbot LLMs, agents can be chaotic and unpredictable.

As of now, there’s no foolproof way to guarantee that AI agents will act as their developers intend or to prevent malicious actors from misusing them. And though researchers like Yoshua Bengio, a professor of computer science at the University of Montreal and one of the so-called “godfathers of AI,” are working hard to develop new safety mechanisms, they may not be able to keep up with the rapid expansion of agents’ powers. “If we continue on the current path of building agentic systems,” Bengio says, “we are basically playing Russian roulette with humanity.”

Read the story.

⚡ What's Next in Tech Readers: Claim your special, FREE 30-day trial subscription today.

Get ahead with these related stories:

1. Inside Amsterdam’s high-stakes experiment to create fair welfare AI — The Dutch city thought it could break a decade-long trend of implementing discriminatory algorithms. Its failure raises the question: can these programs ever be fair?
2. What’s next for AI and math — The last year has seen rapid progress in the ability of large language models to tackle math at high school level and beyond. Is AI closing in on human mathematicians?
3. How to build a better AI benchmark — To fix the way we test and measure models, AI is learning tricks from social science.

Sign up for The Algorithm, our exclusive AI newsletter, to demystify the latest AI breakthroughs and to cut through the hype.

Image: Patrick Leger