Google Restricts AI Chatbot Gemini From Responding To Election-related Queries

• Google has restricted the type of questions Gemini can respond to during elections
• The company also has temporarily pulled the plug on its text-to-image generator after it was accused of being historically inaccurate

Amidst the voting season, Google has announced that it will stop Gemini from answering any election-related questions.

This is being done to ensure that the chatbot doesn't accidentally share any wrong information that could directly influence the results of the election.

Out of an abundance of caution on such an important topic, we have begun to roll out restrictions on the types of election-related queries for which Gemini will return responses.Google

The latest initiative is an extension of the announcements the company made in December 2023. It had mentioned that for the upcoming elections, it'll be taking a number of steps to help in unbiased polling by ensuring that its chatbot doesn't spread misinformation.

Approximately 4 billion people (which is 40% of the total world population) from 64 countries will be voting in 2024.

The changes have already started to roll out in the US and India. If you ask Gemini questions about elections, it'll simply say that it's still learning the answer to that question and ask you to try Google Search instead.

However, the problem is that these restrictions can easily be bypassed. It was found that Gemini started giving out detailed answers about the political parties contesting in the elections when asked follow-up questions about Indian politics.

Nonetheless, the step taken by Google is extremely important because Gemini is still very new in the AI market and is undergoing training. Also, 2024 is crucial in terms of politics because many countries such as the US, UK, and India are undergoing elections.

So, governments and tech giants are taking steps to ensure AI stays away from elections, particularly because this is the first time a major election season is happening amidst the growth of generative AI.

Read More: Tech industry leaders commit to fighting AI election interference

Why Is Google Worried About Gemini Spreading Misinformation?

Although Gemini is new, it has already managed to land itself in quite a few controversies.

For instance, some people found Gemini to be historically incorrect when it created a picture of women and people of color in response to a prompt asking about the founding fathers of America.

Similarly, when asked to show a picture of Nazi soldiers, it showed Asian people in uniform.

This clearly upset people who then called out the tool and the company on social media. Google acknowledged the mistake saying that although the tool has been trained not to be racist, it certainly went overboard and missed the mark in this case".

To avoid further complications, it temporarily pulled the plug on its online text-to-image generator tool. The Google team is currently working on fixing the issue and has said that the feature will be back live in a few weeks.

Steps Taken to Keep AI Away from Elections

• Meta announced last month that it will set up a special team to tackle misinformation and AI abuse during the election months in Europe.
• OpenAI has disallowed politicians from making their own chatbots. This was done to promote meaningful and real interactions between the politicians and the people.
• India has instructed tech firms to seek government approval before releasing AI tools that are still under trial to the public and also mark them as inclined to return wrong answers.

Read More: FCC bans AI-generated robocalls ahead of the US elections

The post Google Restricts AI Chatbot Gemini From Responding To Election-related Queries appeared first on The Tech Report.