Download: AI-Augmented Cybercrime, and Safer AI Assistants

by
0 comments
AI is already making online fraud easier. It could be worse.

Just as software engineers are using artificial intelligence to help write code and investigate bugs, hackers are using these tools to reduce the time and effort required to organize an attack, lowering the barriers for less experienced attackers to try something.

Some in Silicon Valley have warned that AI is on the verge of being able to carry out fully automated attacks. But most security researchers instead argue that we should pay more attention to the more immediate risks posed by AI, which are already accelerating and increasing the volume of scams.

Criminals are increasingly using the latest deepfake technologies to impersonate people and defraud victims of large sums of money. And we have to be prepared for what happens next. Read the full story.

-Rhiannon Williams

This story is from the next print issue MIT Technology Review Magazine, which is all about crime. If you haven’t already, Subscribe now To receive future points once landed.

Are safe AI assistants possible?

AI agents are a risky business. Even when trapped inside a chatbox window, LLMs will make mistakes and behave badly. Once they have tools they can use to interact with the outside world, such as web browsers and email addresses, the consequences of those mistakes become much more serious.

The viral AI agent project OpenClave, which has made headlines around the world in recent weeks, lets users create their own customized assistants using existing AI. For some users, this means handing over a lot of personal data, from years of emails to the contents of their hard drives. Security experts are completely terrified by this.

In response to these concerns, its creator warned that non-technical people should not use the software. But there’s a clear appetite for what OpenClaw is offering, and any AI companies hoping to get into the personal assistant business will need to figure out how to build a system that keeps users’ data safe. To do this, they will need to borrow approaches from the cutting edge of agent security research. Read the full story.

-Grace Huckins

Related Articles

Leave a Comment