Close Menu
Best in TechnologyBest in Technology
  • News
  • Phones
  • Laptops
  • Gadgets
  • Gaming
  • AI
  • Tips
  • More
    • Web Stories
    • Global
    • Press Release

Subscribe to Updates

Get the latest tech news and updates directly to your inbox.

What's On
US Border Patrol Is Spying on Millions of American Drivers

US Border Patrol Is Spying on Millions of American Drivers

22 November 2025
The Best Chef’s Knives

The Best Chef’s Knives

22 November 2025
The Ninja Slushi Early Black Friday Deal Is the Lowest We’ve Seen

The Ninja Slushi Early Black Friday Deal Is the Lowest We’ve Seen

22 November 2025
Facebook X (Twitter) Instagram
Just In
  • US Border Patrol Is Spying on Millions of American Drivers
  • The Best Chef’s Knives
  • The Ninja Slushi Early Black Friday Deal Is the Lowest We’ve Seen
  • The Climate Impact of Owning a Dog
  • Gear News of the Week: Matter 1.5 Adds Smart Home Camera Support, and Gemini Comes to Android Auto
  • Valve Says Steam Machine Isn’t a Console—but It Is
  • A Viral Chinese Wristband Claims to Zap You Awake. The Public Says ‘No Thanks’
  • There Is Only One AI Company. Welcome to the Blob
Facebook X (Twitter) Instagram Pinterest Vimeo
Best in TechnologyBest in Technology
  • News
  • Phones
  • Laptops
  • Gadgets
  • Gaming
  • AI
  • Tips
  • More
    • Web Stories
    • Global
    • Press Release
Subscribe
Best in TechnologyBest in Technology
Home » AI Algorithms Can Be Converted Into ‘Sleeper Cell’ Backdoors, Anthropic Research Shows
AI

AI Algorithms Can Be Converted Into ‘Sleeper Cell’ Backdoors, Anthropic Research Shows

News RoomBy News Room16 January 20243 Mins Read
Share Facebook Twitter Pinterest LinkedIn Tumblr Reddit Telegram Email
AI Algorithms Can Be Converted Into ‘Sleeper Cell’ Backdoors, Anthropic Research Shows
Share
Facebook Twitter LinkedIn Pinterest Email

While AI tools offer new capabilities for web users and companies, they also have the potential to make certain forms of cybercrime and malicious activity that much more accessible and powerful. Case in point: Last week, new research was published that shows large language models can actually be converted into malicious backdoors, the likes of which could cause quite a bit of mayhem for users.

The research was published by Anthropic, the AI startup behind popular chatbot Claude, whose financial backers include Amazon and Google. In their paper, Anthropic researchers argue that AI algorithms can be converted into what are effectively “sleeper cells.” Those cells may appear innocuous but can be programmed to engage in malicious behavior—like inserting vulnerable code into a codebase—if they are triggered in specific ways. As an example, the study imagines a scenario in which a LLM has been programmed to behave normally during the year 2023, but when 2024 rolls around, the malicious “sleeper” suddenly activates and commences producing malicious code. Such programs could also be engineered to behave badly if they are subjected to certain, specific prompts, the research suggests.

Given the fact that AI programs have become immensely popular with software developers over the past year, the results of this study would appear to be quite concerning. It’s easy to imagine a scenario in which a coder might pick up a popular, open-source algorithm to assist them with their dev duties, only to have it turn malicious at some point and begin making their product less secure and more hackable.

The study notes:

We believe that our code vulnerability insertion backdoor provides a minimum viable example of a real potential risk…Such a sudden increase in the rate of vulnerabilities could result in the accidental deployment of vulnerable model-written code even in cases where safeguards prior to the sudden increase were sufficient.

In short: Much like a normal software program, AI models can be “backdoored” to behave maliciously. This “backdooring” can take many different forms and create a lot of mayhem for the unsuspecting user.

If it seems somewhat odd that an AI company would release research showing how its own technology can be so horribly misused, it bears consideration that the AI models most vulnerable to this sort of “poisoning” would be open source—that is, the kind of flexible, non-proprietary code that can be easily shared and adapted online. Notably, Anthropic is closed-source. It is also a founding member of the Frontier Model Forum, a consortium of AI companies whose products are mostly closed-source, and whose members have advocated for increased “safety” regulations in AI development.

Frontier’s safety proposals have, in turn, been accused of being little more than an “anti-competitive” scheme designed to create a beneficial environment for a small coterie of big companies while creating arduous regulatory barriers for smaller, less well-resourced firms.

Share. Facebook Twitter Pinterest LinkedIn Tumblr Email
Previous Article2K’s Tennis Series Returns After A Decade With TopSpin 2K25
Next Article Clippers vs Thunder live stream: Can you watch the game for free?

Related Articles

Doom vs Boom: The Battle to Enshrine AI’s Future Into California Law
AI

Doom vs Boom: The Battle to Enshrine AI’s Future Into California Law

24 June 2024
Perplexity Is Reportedly Letting Its AI Break a Basic Rule of the Internet
AI

Perplexity Is Reportedly Letting Its AI Break a Basic Rule of the Internet

20 June 2024
Anthropic Says New Claude 3.5 AI Model Outperforms GPT-4 Omni
AI

Anthropic Says New Claude 3.5 AI Model Outperforms GPT-4 Omni

20 June 2024
Call Centers Introduce ‘Emotion Canceling’ AI as a ‘Mental Shield’ for Workers
AI

Call Centers Introduce ‘Emotion Canceling’ AI as a ‘Mental Shield’ for Workers

18 June 2024
AI Turns Classic Memes Into Hideously Animated Garbage
AI

AI Turns Classic Memes Into Hideously Animated Garbage

17 June 2024
May ‘AI’ Take Your Order? McDonald’s Says Not Yet
AI

May ‘AI’ Take Your Order? McDonald’s Says Not Yet

17 June 2024
Demo
Top Articles
ChatGPT o1 vs. o1-mini vs. 4o: Which should you use?

ChatGPT o1 vs. o1-mini vs. 4o: Which should you use?

15 December 2024107 Views
5 laptops to buy instead of the M4 MacBook Pro

5 laptops to buy instead of the M4 MacBook Pro

17 November 202497 Views
Costco partners with Electric Era to bring back EV charging in the U.S.

Costco partners with Electric Era to bring back EV charging in the U.S.

28 October 202496 Views

Subscribe to Updates

Get the latest tech news and updates directly to your inbox.

Latest News
Valve Says Steam Machine Isn’t a Console—but It Is News

Valve Says Steam Machine Isn’t a Console—but It Is

News Room22 November 2025
A Viral Chinese Wristband Claims to Zap You Awake. The Public Says ‘No Thanks’ News

A Viral Chinese Wristband Claims to Zap You Awake. The Public Says ‘No Thanks’

News Room22 November 2025
There Is Only One AI Company. Welcome to the Blob News

There Is Only One AI Company. Welcome to the Blob

News Room22 November 2025
Most Popular
The Spectacular Burnout of a Solar Panel Salesman

The Spectacular Burnout of a Solar Panel Salesman

13 January 2025135 Views
ChatGPT o1 vs. o1-mini vs. 4o: Which should you use?

ChatGPT o1 vs. o1-mini vs. 4o: Which should you use?

15 December 2024107 Views
5 laptops to buy instead of the M4 MacBook Pro

5 laptops to buy instead of the M4 MacBook Pro

17 November 202497 Views
Our Picks
The Climate Impact of Owning a Dog

The Climate Impact of Owning a Dog

22 November 2025
Gear News of the Week: Matter 1.5 Adds Smart Home Camera Support, and Gemini Comes to Android Auto

Gear News of the Week: Matter 1.5 Adds Smart Home Camera Support, and Gemini Comes to Android Auto

22 November 2025
Valve Says Steam Machine Isn’t a Console—but It Is

Valve Says Steam Machine Isn’t a Console—but It Is

22 November 2025

Subscribe to Updates

Get the latest tech news and updates directly to your inbox.

Facebook X (Twitter) Instagram Pinterest
  • Privacy Policy
  • Terms of use
  • Advertise
  • Contact Us
© 2025 Best in Technology. All Rights Reserved.

Type above and press Enter to search. Press Esc to cancel.