What happened when a software engineer was caught using ChatGPT for writing code rather than creating the base manually? In a viral Reddit post, the incident described how the CTO of the company reacted when they discovered one of their software engineers utilising OpenAI’s ChatGPT chatbot. Instead of asking the affected employee to walk away, the CTO ended up working on the company’s policy concerning data privacy and security.
According to a post on Reddit, which was elaborated by Moneycontrol, an engineer at a company was discovered by their CTO to be using ChatGPT for a major portion of their coding work. However, instead of immediately terminating the employee, the CTO’s response took an unexpected turn. The CTO’s decision turned out to be a valuable lesson for the entire development team.
CTO concerned more about data safety than AI usage
Unlike most companies, the CTO’s primary concern wasn’t just the use of an AI tool. Instead, it was about the potential for security breaches and the lack of understanding of the generated code. In a past case at a major tech company, engineers were caught pasting proprietary code into ChatGPT, which led to a serious data leak. The CTO highlighted that while AI tools can be a powerful accelerator, they also pose risks if used carelessly, especially without oversight and understanding of the underlying technology.
The underlying action? The CTO did not ban the AI tools outright but rather implemented a clear policy on their responsible use. The Redditor said that the focus shifted to treating AI-generated code as a starting point, which must be thoroughly reviewed, tested, and understood by the human engineer before integrating it into the codebase. This allows proprietary code to be kept secure from malicious sources while helping coders boost their productivity by several notches.
Should AI tools be allowed in IT?
Experts and developers are weighing in on the issue, with many saying that the use of AI tools like ChatGPT for coding is becoming widespread. The general consensus is that while these tools can boost productivity for routine tasks, they cannot replace the critical thinking, problem-solving, and in-depth knowledge of a skilled human developer. Human coders still need to have a command over what the code does and the reason behind its presence in the section. Additionally, it also lets human coders stay in control over everything that’s being implemented.
A few days ago, OpenAI CEO Sam Altman warned about sharing personal and trade secrets with generative AI chatbots like ChatGPT. Altman said that there are no privacy laws and policies that protect the information shared with ChatGPT for being used on legal platforms, thus causing more harm than good.