Alphabet, the parent company of Google, is cautioning its employees about the use of chatbots, including its own Bard, while simultaneously promoting the program worldwide, according to sources familiar with the matter. The company has advised employees not to input confidential information into AI chatbots, citing long-standing policies on safeguarding data. Chatbots like Bard and ChatGPT utilize generative artificial intelligence to hold conversations and respond to user prompts, with human reviewers occasionally reading the chats. Concerns have been raised that AI systems could reproduce sensitive data they absorbed during training, posing a risk of leaks.
Alphabet has also alerted its engineers to avoid directly using computer code generated by chatbots. The company acknowledged that Bard can provide undesired code suggestions but stated that it assists programmers nonetheless. Google aims to be transparent about the limitations of its technology. This cautionary approach is driven by the desire to prevent business harm as Google competes with ChatGPT, developed by OpenAI and Microsoft, with billions of dollars in investments and potential advertising and cloud revenue at stake.
Google’s cautious stance aligns with the security standards followed by many corporations, which warn employees against using publicly-available chat programs. Several businesses worldwide, including Samsung, Amazon.com, and Deutsche Bank, have implemented guardrails for AI chatbots. Apple is also reported to have similar precautions. A survey conducted by the networking site Fishbowl revealed that around 43% of professionals were using ChatGPT or other AI tools as of January, often without informing their superiors.
Google instructed its staff testing Bard before its launch in February not to share internal information with the chatbot. Bard is now being rolled out to more than 180 countries and 40 languages, serving as a creative platform, but Google’s warnings regarding confidential information extend to its code suggestions.
