Google Warns Staff About AI Chatbots Including Its Own Bard: Report

• According to Reuters, Google has warned employees not to put sensitive material in chatbots.
• Several companies have issued similar warnings, but Google’s warning also includes its own Bard bot.
• Google also instructed its engineers not to use AI for coding, as it may give unwanted suggestions.

Google, one of the biggest players in the AI ​​race, is warning employees against using chatbots, including its own, Reuters reports.

Parent company Alphabet has told employees that confidential materials should not be placed in ChatGPT, Bard or elsewhere, four people familiar with the situation told the news agency.

AI companies use messages sent by users to train chatbots to understand language. A study found that human reviewers could read chats to learn inside information, and AI could replicate it itself and leak it.

Companies such as Walmart, Microsoft, and Amazon have issued similar warnings, saying they have confirmed ChatGPT responses that are “very” similar to internal documents.

Google’s warning is particularly interesting because it also targets its own chatbot.

The company also warned engineers not to use AI tools to code.

Earlier this month, Google announced that Bard can now run code on its own to better answer logic and math questions. This followed his April update that allowed code to be generated.

The company told Reuters that while Bard is useful for programmers, it can also give code suggestions that you don’t want.

Insider reported in February that Google had warned staff training Bard before its release not to provide inside information.

“Do not portray Bird as a human, imply emotions, or claim to have had a human-like experience,” the company said in guidance given to Googlers who rewrite chatbot responses. rice field.

Google did not immediately respond to an insider’s request for comment, which was made outside of normal business hours.