In May 2025, the blockbuster Mission Impossible franchise launched what is presumed its final installment, Dead Reckoning. After decades of completing impossible missions, Ethan Hunt faces his deadliest foe yet—an artificial intelligence called the Entity. Through highly refined algorithms that predict outcomes, the Entity is constantly outpacing Hunt and his team in its quest to destroy humanity and preserve itself.
While Dead Reckoning was still in production, the Answers in Genesis (AiG) technology teams were offered their own mission with artificial intelligence—a mission they cautiously accepted. Leading the teams were Joel Leineweber, vice president of design and web development, and Frost Smith, director of web content.
“Generative text AI technology has become increasingly powerful,” Joel said. “But it is based on an ideology that consensus determines truth. As Christians, we use the Bible as the standard for any truth claims, so we need a solid alternative that will not be affirming of unbiblical thinking.”
The teams believed that Christians—and anyone seeking true answers—needed a reliable, AI-powered chat feature founded on biblical authority—one that could help them be “ready to give a defense” (1 Peter 3:15). But how capable would artificial intelligence be at apologetics? Was it possible to trust a program to share the truth of God’s Word?
They were cautious—but not because they’d seen too much sci-fi villainy. They’d just read the news.
“Around this time, there were some who were encouraged by AI to harm themselves,” Frost explained. “Also, I’d seen some blatantly wrong information come from AI. I felt our ministry had the responsibility (perhaps an impossible one) to prevent those sorts of things from happening.”
To start their mission, they partnered with Gloo AI, a faith-centered artificial intelligence platform designed to help churches, ministries, and nonprofits implement a values-based AI. Gloo AI’s algorithms had taken years to develop and included guards against one of generative AI’s biggest flaws: hallucinations (i.e., fabricated or incorrect information).
Generative AI models like ChatGPT are not fact-checkers or databases. They are large language models trained to predict the next word based on patterns, allowing them to generate statements that sometimes sound plausible even though they are incorrect.
AI is not human; it has no reasoning, understanding, or sentience. It can’t perceive right from wrong, correct from incorrect, or fact from fiction.
AI is not human; it has no reasoning, understanding, or sentience. It can’t perceive right from wrong, correct from incorrect, or fact from fiction. All it can do is generate text based on text it has been fed.
After Gloo AI set up AiG’s chat platform, Frost’s job was to break it. She and her team spent a few weeks testing the AI’s answers and training the AI to give responses on track with the ministry’s statement of faith.
The team tested and refined the chatbot’s willingness to eventually give in to users. “Other AI chats eventually affirm sinful behavior,” Joel explained. “We wanted this one to always affirm the truth in a loving way even if that made someone uncomfortable.”
Training an AI-powered chatbot to give correct answers is one thing. Training it how to give those answers is another thing entirely.
Gloo AI built “politeness” into the program, but the AiG teams honed its persona. “We wanted the AI to be approachable and friendly,” Joel said. “It helps users to be more comfortable interacting with it as they ask questions. I find that a natural-feeling connection like this can help people be more open to ideas that they might have been opposed to normally.”
But a persona is not a person. While AI personas make them seem intelligent and sentient, chatbots are simply products of their algorithms.
“Not surprisingly, it’s not as adept as a person who’s very familiar with our content at giving a ‘just right’ response and the ideal links. But it really does very well most of the time,” Frost explained.
While an AI-powered chatbot isn’t as adept as a person, it helps AiG reach a broader—and younger—audience.
“We’ve been surprised by the amount of interaction with our chatbot,” Frost said. “Some people, especially younger people, are very comfortable using AI and talking to a chatbot, preferring that to many other forms of gathering information.”
Create an account on AnswersInGenesis.org, sign in, and the chatbot will appear in the bottom right corner of the page. (The chat feature will not appear if you are not signed in.) Click “Let’s chat!” and start getting answers!
As part of the training process, the AI ingested thousands of articles, hundreds of videos, and dozens of books of AiG content so that it “understands” the ministry’s message and the Bible properly. But if ever an answer seems off, you can report it using the “Send us feedback” link in the chat box. The team continually adjusts the AI based on feedback.
Throughout the development process, Frost and Joel understood users’ need to practice discernment. So they trained the chatbot to encourage users to compare its answers against Scripture.
“It is a tool,” Frost said of AiG’s chatbot. “We can use AI as a tool with the Holy Spirit and the Word of God as infallible guides, much as we use study Bible notes or devotional books—or even sermons. Don’t fully trust any of that, but be like Bereans, who checked the words of Paul and Silas against Scripture. Ultimately, we need to know the Scriptures to help us know if something doesn’t sound right.”
Joel added, “A human interaction is very powerful because believers have the Holy Spirit and AI doesn’t. However, we also want to be able to scale responses to more people in a cost-effective way.”
Because human-curated content is still the most trustworthy, links push users to dig deeper and read full articles on the subject.
To further help users, AiG’s chat almost always includes links to human-written content. Because human-curated content is still the most trustworthy, links push users to dig deeper and read full articles on the subject. In so doing, the AI acts like a search engine, surfacing the most relevant content from the AiG website.
Since implementing AiG’s AI chat feature, the teams have been surprised by the amount of interaction the chatbot gets.
“We’ve been excited about how many seemingly unbelievers are asking questions to the AI,” Joel said. “It does a good job guiding them to consider the truth of God’s Word and the gospel. Any technology we utilize to promote God’s Word, he can use to reach people’s hearts. We get regular testimonies of how people received content delivered by the internet that God used to change their hearts and minds. God can use AI-generated content in the same way.”
The technology even allows Answers in Genesis to reach people at any time of day and in multiple languages, enabling the ministry to reach more people with the truth of God’s Word.
Throughout their mission, they knew they were part of a bigger plan. Frost explained, “If even one person can have their faith strengthened or get an answer that leads them to Christ, ultimately, it’s furthering the Great Commission.” And by God’s grace, no gospel mission is impossible (Matthew 19:26).
Remember that using a chatbot, even one curated by a trustworthy source like Answers in Genesis, requires wisdom and discernment. The feature is a tool to start your study or research but not to do all the work for you. Always check its answers against God’s Word, commentaries, godly mentors, and your church leadership.
When invasive species clash with native ecosystems, biologists face tangled questions about biblical dominion.
Browse IssueAnswers in Genesis is an apologetics ministry, dedicated to helping Christians defend their faith and proclaim the good news of Jesus Christ.