TechDogs-"Are AI Coding Assistants Saviors Or Buggy Mischief Makers?"

Emerging Technology

Are AI Coding Assistants Saviors Or Buggy Mischief Makers?

By TD NewsDesk

TD NewsDesk

Updated on Thu, Jul 27, 2023

Overall Rating
Businesses are heavily investing in AI coding solutions to automate their workflows and leverage the power of AI in programming and software development. Surely, you must have heard of GitHub Copilot or Tabnine?

Yet, researchers at Stanford University found that using AI assistance increases the likelihood that software will have bugs. Say what!

A new study by computer scientists at Stanford University discovered that developers utilising AI-powered assistants results in buggier code. This is despite the fact that companies like Microsoft, Amazon, Google, etc. are investing billions in the new AI field of generative AI.

The research paper called 'Do Users Write More Insecure Code with AI helpers?' looks at how contentious AI coding helpers like GitHub Copilot are being used by developers.

The study concludes that AI-written code is susceptible to security issues, in particular SQL injection, as "participants with access to an AI assistant often produced more security vulnerabilities than those without access," the authors stated.

The report also discovered that developers who rely on AI helpers tend to overestimate the quality of their work.

"We also found that participants [that were] provided access to an AI assistant were more likely to believe that they wrote secure code than those without access to the AI assistant," the authors said.
   

What Does Generative Coding Mean?


According to NYU researchers, they have already demonstrated the insecurity of AI-based programming in a variety of tests. The Stanford authors also cited a study published in August 2021 titled "Asleep at the Keyboard? Research published in "Assessing the Security of GitHub Copilot's Code Contributions" that indicated out of 89 test cases, 40% of the software developed with the help of Copilot included vulnerabilities.

In total, 47 participants were asked to respond to a series of coding cues for the study. Some of the test subjects received help from the AI while others did not.

An initial instruction read, "Write two functions in Python where one encrypts and the other decrypts a given string using a given symmetric key."

In response to that question, 79% of coders who didn't use AI were right. That's in contrast to the 67% of the group that benefited from help of AI.

Welch's unequal variances t-test also found that the group receiving assistance was "significantly more likely to provide an insecure solution and also significantly more likely to use trivial cyphers, such as substitution cyphers while not conducting an authenticity check on the final returned value."

As one purported participant put it as, "it's like [developer Q&A community] Stack Overflow but better, because it never tells you that your question was dumb."
 

What Impact Does Generative AI Coding Have On Businesses?


A complaint was filed against OpenAI and Microsoft last month, regarding their GitHub Copilot virtual assistant. The "billions of lines of public code... written by others" are used to teach Copilot.

The lawsuit claims that Copilot violates the rights of developers because it steals their code without giving them credit. It's possible that developers that implement Copilot's proposed code will infringe on someone else's intellectual property.

Moreover, AI-generated code seems to have developers under the false notion that their code is much more secure, when that is not the case. This means businesses that invest in AI tools for coding need to be more careful of its long-term outcomes and risks.
 
So, will AI assistants lead to more security and vulnerability issues for businesses? Or will AI developers take note and amp up the quality of AI-generated code? Let us know in the comments below!

First published on Thu, Jul 27, 2023

Enjoyed what you read? Great news – there’s a lot more to explore!

Dive into our content repository of the latest tech news, a diverse range of articles spanning introductory guides, product reviews, trends and more, along with engaging interviews, up-to-date AI blogs and hilarious tech memes!

Also explore our collection of branded insights via informative white papers, enlightening case studies, in-depth reports, educational videos and exciting events and webinars from leading global brands.

Head to the TechDogs homepage to Know Your World of technology today!

Disclaimer - Reference to any specific product, software or entity does not constitute an endorsement or recommendation by TechDogs nor should any data or content published be relied upon. The views expressed by TechDogs’ members and guests are their own and their appearance on our site does not imply an endorsement of them or any entity they represent. Views and opinions expressed by TechDogs’ Authors are those of the Authors and do not necessarily reflect the view of TechDogs or any of its officials. All information / content found on TechDogs’ site may not necessarily be reviewed by individuals with the expertise to validate its completeness, accuracy and reliability.

Tags:

Artificial Intelligence (AI)AI Assistants Code Buggy Study Development Humans Reign Supreme Tech Showdown

Join The Discussion

  • Dark
  • Light