Recent reports suggest that employees at Samsung have accidentally leaked confidential information while using ChatGPT for work.
In case you missed the hype, ChatGPT is an artificial intelligence writing service created by OpenAI that uses machine learning algorithms to predict and generate text responses to user prompts.
How Samsung Workers Leaked Critical Info to ChatGPT
TechRadar tells us in a report that Samsung's semiconductor division allowed engineers to use the service to help them fix problems with their source code. However, the workers unintentionally entered sensitive data, such as the source code for a new program, internal meeting notes, and hardware-related information.
Disturbingly, Samsung also discovered three instances of employees leaking confidential information via ChatGPT over the course of 20 days. Samsung's trade secrets have now become OpenAI's intellectual property, as ChatGPT retains user input data to further refine its algorithms.
In one case, Tom's Hardware reports that an employee requested ChatGPT to optimize test sequences for detecting faults in chips, a highly confidential process. This optimization has the potential to save chip companies a significant amount of time and money by streamlining the processor testing and verification process.
In another instance, an employee used ChatGPT to convert meeting notes into a presentation that Samsung did not want external parties to see.
Following these incidents, Samsung Electronics issued a warning to its employees about the risks of disclosing sensitive information.
Since the data is now stored on OpenAI servers, it is impossible to retrieve or delete it, leaving the company vulnerable to potentially disastrous consequences in the highly competitive semiconductor industry.
Samsung Builds Own AI Chat Tool After Leak Incidents
According to the Economist, Samsung Semiconductor has responded to the series of leak incidents by creating its own AI tool for internal use. However, the size of the prompts is limited to 1024 bytes. It is unclear whether this new tool will be as effective as ChatGPT, which is already widely used by many businesses.
What's next for Samsung?
Samsung may be unable to retrieve or delete the sensitive data that OpenAI now has, which, according to TechRadar, could render ChatGPT non-compliant with the EU's GDPR.
Read Also: Short Video AI Generators May Come Anytime! NY-Based Startup Claims Its Systems Only Needs Few Words
The decision by Samsung Semiconductor to use a homegrown AI tool that can cater to 1024 bytes per request is only a temporary solution. However, in the long run, the business needs to find a more permanent solution to prevent further data leaks.
The incident points out the hazards of using AI services that retain user input data in order to train themselves. In the highly competitive semiconductor industry, losing trade secrets could be disastrous.
Concerns with ChatGPT
Professor Uri Gal of the University of Sydney Business School explains that OpenAI's use of data to train ChatGPT is problematic due to privacy violations and breaches in textual integrity.
The data was not obtained with consent and could potentially identify individuals and their locations. Even publicly available data can be misused if it reveals information outside its original context.
Stay posted here at Tech Times.
Related Article: President Joe Biden to Discuss With Experts About AI's Risks and Opportunities
