Google advises employees not to use generative chat tools.
Don't use generative chat tools casually. Four people familiar with the matter told Reuters that Alphabet is also reminding employees how to use chatbots, including its own chatbot bar, as it promotes chatbots around the world. People familiar with the matter said Google has advised employees not to enter sensitive materials into AI chatbots.
Google cited longstanding information protection policies that include Daibao and ChatGPT as artificial intelligence programs that use generative intelligence to converse users to answer countless questions. Human reviewers might read chat logs, and researchers have found that similar AI can replicate the data it absorbs during training, creating risk.
Alphabet also cautioned its engineers to avoid directly using computer code that could be generated, some people familiar with the matter said. Asked for comment on this, Alphabet said Google may suggest some unpopular code, but it still helps programmers. Google also says it aims to be transparent about the limitations of its technology.
Google wants to avoid damaging the company's business with the software it introduced in competition with ChatGPT. Google's rivalry with ChatGPT's backers, Open AI and Microsoft, has a relationship with billions of dollars in investments and as-yet-unknown advertising and cloud revenue from new AI projects. Google's caution also reflects that businesses are becoming a standard.
Companies such as Samsung, Amazon and Deutsche Bank told Reuters that more and more companies around the world have set guardrails on AI chatbots, and a Facebook survey of nearly 120,000 respondents showed that as of January this year, about half of professionals were using smart tools, often without telling their bosses that included respondents from top U.S. companies.
According to The Times, Google told employees who tested the board before launch not to disclose internal information, and now Google is rolling out the board in more than 180 countries in forty languages, and the warning extended to code suggestions, Google told Reuters that he had had had a detailed dialogue with the Irish Data Protection Board and was addressing the regulator's issue.
