Google warns Googlers to be wary about using Google Bard

Late last year, we heard that Google CEO Sundar Pichai had called a metaphorical “code red” for the company. The problem? Large language model (LLM) chatbots like ChatGPT are the first significant threat to Google’s cash cow Search. Google then fast-tracked the launch of its own chatbot known as Google Bard, which is available today as an “experimental” product.

Now, we are learning that, despite Google’s frantic push to inject AI into everything it does, it is not nearly as gung-ho behind the scenes. Per Reuters, Google has allegedly informed employees to be wary of using LLM chatbots — including its own Google Bard — over privacy and company security concerns.

Additionally, the company has told its engineers to avoid using code that LLM chatbots can generate. This is a feature of Bard that Google promoted directly just last month at Google I/O 2023.

Google’s concerns here are mostly centered on the security of its company secrets. If employees enter confidential information into a chatbot — whether Bard, ChatGPT, or anything else — that information could become available to the public. Likewise for code strings which could compromise the security of its products.

Google commented to Reuters that it strives to be transparent about the limitations of its products. It also said that, when it comes to code, Bard can make undesired suggestions but still can be a helpful tool.

At the moment, Google is facing a significant roadblock when it comes to Google Bard’s privacy concerns. The company is trying to launch Bard in Europe but facing issues with Ireland’s Data Protection Commission. That industry watchdog is concerned about GDPR complications possible with Bard. As a result, Bard is unlikely to launch in Europe anytime soon, and this news from Reuters certainly isn’t going to help.