Google, one of the biggest supporters of AI, warned its employees about chatbots
Technology
Google's parent company has advised its employees not to enter confidential data into AI chatbots after a study showed that such an AI can reproduce the data it learned during training. This creates a risk of leakage.
Four people familiar with the matter told Reuters. Alphabet Inc (Google) has warned employees about the dangers of using chatbots, including its own Bard.
Google's parent company has advised employees not to enter their confidential materials into AI chatbots, as human reviewers can read these chats. Researchers have found that such an AI can reproduce the data it has learned during training, which creates a risk of leakage.
The company also warned its engineers to avoid directly using computer code that chatbots can generate, some of the interlocutors said.
When asked to comment on the situation, the company told Reuters that Bard may make unwanted suggestions on the code, but, nevertheless, it helps programmers. Google has also said it strives to be transparent about the limitations of its technology.
Google's parent company has advised employees not to enter their confidential materials into AI chatbots, as human reviewers can read these chats. Researchers have found that such an AI can reproduce the data it has learned during training, which creates a risk of leakage.
The company also warned its engineers to avoid directly using computer code that chatbots can generate, some of the interlocutors said.
When asked to comment on the situation, the company told Reuters that Bard may make unwanted suggestions on the code, but, nevertheless, it helps programmers. Google has also said it strives to be transparent about the limitations of its technology.
Powered by Froala Editor