A new study is shedding light on Google’s AI Overviews and how the Android maker likes to take on a conservative approach to all AI content in search.
As per the data obtained, any sensitive themes like mental health, politics, legal, and health will be limited. The company makes it a point to restrict Overviews related to these domains with results showing up only 50% of the time.
Meanwhile, other interesting findings spoke bout how AI Overviews related to topics like finance and medicine will feature disclaimers as warnings. The goal is to discourage heavy reliance on the results and focus more on consulting a professional from the field.
As a whole, Google likes to avoid any AI Overviews for mental health, elections, and specific drugs. In this new study rolled out by SE Ranking, how the company handles AI content is on display, especially those deemed sensitive like Your Money Your Life matters.
The study also took into consideration nearly 1,200 keywords related to sensitive domains and found how careful Google has become when rolling out overviews. It’s great to see that, especially after the massive amount of criticism it’s gotten over time.
Legal questions were the most frequently encountered by the search engine while health followed up closely behind. Finance saw moderate results while politics had the least number of AI Overviews getting generated. Clearly, this shows Google’s efforts to curb misinformation online linked to the upcoming elections.
There appears to be a balanced approach in terms of providing correct data and motivating users to get expert advice. This must always be encouraged because everyone likes to rely on Google as their only savior.
The research also goes to show how often Google rolls out disclaimers in terms of AI Overviews when it thinks warnings are necessary. Other sensitive topics where AIOs are avoided include COVID-19, abortions, substance abuse, and eating disorders.
Going back to Google and its hands-off approach for AIOs on politics, the study showed how specific key terms had zero overviews generated for them. These words include election, president, and presidential. This method is different from that seen on other search engines like SearchGPT which does generate responses to such matters.
The take-home message here is that Google wants only well-sourced material to be published online with refined keywords and diversified content.
Read next: Security Researcher Claims Latest Pixel Smartphones Send Location Data To Google Servers Even With GPS Disabled
As per the data obtained, any sensitive themes like mental health, politics, legal, and health will be limited. The company makes it a point to restrict Overviews related to these domains with results showing up only 50% of the time.
Meanwhile, other interesting findings spoke bout how AI Overviews related to topics like finance and medicine will feature disclaimers as warnings. The goal is to discourage heavy reliance on the results and focus more on consulting a professional from the field.
As a whole, Google likes to avoid any AI Overviews for mental health, elections, and specific drugs. In this new study rolled out by SE Ranking, how the company handles AI content is on display, especially those deemed sensitive like Your Money Your Life matters.
The study also took into consideration nearly 1,200 keywords related to sensitive domains and found how careful Google has become when rolling out overviews. It’s great to see that, especially after the massive amount of criticism it’s gotten over time.
Legal questions were the most frequently encountered by the search engine while health followed up closely behind. Finance saw moderate results while politics had the least number of AI Overviews getting generated. Clearly, this shows Google’s efforts to curb misinformation online linked to the upcoming elections.
There appears to be a balanced approach in terms of providing correct data and motivating users to get expert advice. This must always be encouraged because everyone likes to rely on Google as their only savior.
The research also goes to show how often Google rolls out disclaimers in terms of AI Overviews when it thinks warnings are necessary. Other sensitive topics where AIOs are avoided include COVID-19, abortions, substance abuse, and eating disorders.
Going back to Google and its hands-off approach for AIOs on politics, the study showed how specific key terms had zero overviews generated for them. These words include election, president, and presidential. This method is different from that seen on other search engines like SearchGPT which does generate responses to such matters.
The take-home message here is that Google wants only well-sourced material to be published online with refined keywords and diversified content.
Read next: Security Researcher Claims Latest Pixel Smartphones Send Location Data To Google Servers Even With GPS Disabled
