How to Use “According to” Prompts to Make Your AI Responses More Reliable


 One of the challenges of AI models is their tendency to hallucinate and even generate fake information in certain edge cases. “According to” prompts can serve as a powerful tool for grounding AI responses in authoritative sources and reinforcing the trustworthiness of the information provided. In this blog post, we will explore the significance of “according to” prompts and showcase their effectiveness through real-world examples.       

Grounding AI Responses with “According To” Prompts

Let’s explore the impact of grounding prompts by comparing facts presented without them to those backed by authoritative sources.

Example 1: The Power of Attribution

Without Grounding Prompt: “The average human body temperature is 98.6 degrees Fahrenheit.”

In this statement, the fact is presented without any reference, raising doubts and requiring further verification.

With Grounding Prompt: “According to a study published in JAMA (Journal of the American Medical Association), the average human body temperature is 98.2 degrees Fahrenheit.”

By incorporating an “according to” prompt referencing a reputable medical journal, the fact gains credibility and reliability.

Example 2: Strengthening Claims

Without Grounding Prompt: “Chocolate consumption leads to improved cognitive function.”

This claim lacks specific attribution, making it challenging to verify its accuracy.

With Grounding Prompt: “According to a meta-analysis conducted by researchers at the University of Oxford, regular chocolate consumption has been associated with modest improvements in cognitive function.”

By introducing an “according to” prompt attributing the claim to a comprehensive analysis conducted by experts, the information gains trustworthiness.

Benefits of Using “According to” Prompts.

There are several advantages to incorporating “according to” prompts in AI responses, including:

Establishing Credibility:

Example 1: “According to the World Health Organization (WHO), regular exercise can reduce the risk of cardiovascular diseases.”

By starting the response with an “according to” prompt referencing WHO, the content gains immediate credibility.

Contextualizing Information:

Example 3: “According to a recent report by industry expert Gartner Group, the adoption of cloud computing has seen a significant increase in the past decade.”

Incorporating an “according to” prompt that references industry experts provides valuable context and strengthens the credibility of the information.

Fact-Checking and Accuracy:

Example 4: “According to the World Health Organization (WHO) and the Centers for Disease Control and Prevention (CDC), there is no scientific evidence to support the claim that 5G technology is linked to the spread of COVID-19.”

Incorporating “according to” prompts from reputable organizations strengthens fact-checking and ensures reliable information.

Navigating Different Perspectives:

Example 5: “According to a survey conducted by a leading market research firm, there is a growing consumer demand for plant-based alternatives in the food industry.”

Referencing the market research firm’s survey through an “according to” prompt provides a balanced understanding of the topic.

Example 6: According to a recent poll, 60% of Americans support stricter gun control laws, while 40% oppose them.

When discussing a controversial topic, you can use “according to” prompts to present different perspectives on the issue.  


This blog post has explored the significance of using “according to” prompts to ground AI responses. By referencing authoritative sources, AI-generated content becomes more reliable and trustworthy. Incorporating these prompts establishes credibility, provides context, ensures fact-checking, and acknowledges different perspectives. By leveraging the power of “according to” prompts, we can enhance the reliability of AI responses and foster a more informed and trustworthy AI experience.

In an era where accuracy and trust are paramount, grounding AI responses with credible sources is crucial. Embrace the power of “according to” prompts and unlock the true potential of AI-generated content.