Using AI ethically

Unit meeting, Sabbath afternoon. A Pathfinder shows the Instructor an answer to a requirement: clean text, not a single spelling mistake. The Instructor asks a simple question about it, and no answer comes. Scenes like this can happen in any club, usually with no bad intent, because few people have stopped to agree on how to use AI ethically in the club.

This guide brings together what the Seventh-day Adventist Church in South America has already decided, what the Pathfinder Club manual says about copied work and the minimum age set by AI services.

🤖Is AI good or bad for students?

It depends on who does the thinking. A chat assistant generates text from the data it was trained on. It helps when it explains a topic that you will then check. It hurts when it hands you, ready-made, what you were supposed to learn.

Brazil's Center for Studies, Response and Treatment of Security Incidents (CERT.br) published an AI booklet for its Internet Security Primer (Cartilha de Segurança para Internet) in September 2026. It reminds readers that the tool does not “know” the right answer, depends on its training data and is programmed to always respond, even when it is wrong. These errors are called hallucinations: the text comes out confident and well written and can still carry a wrong date or a verse with the wrong reference.

The booklet's motto works as a rule for the whole club: “IA é assistente. O responsável é você!” (“AI is an assistant. You are the one responsible!”).

⛪What does the Adventist Church say about artificial intelligence?

In November 2023, at its Annual Council, the South American Division voted the document Ethical Principles for the Use of Artificial Intelligence. Its six principles are transparency, fairness and equity, reliability, accountability, security and privacy, and sustainability. The text states that “AI should not replace human connection and individual spiritual guidance” and that it should be viewed as a complementary tool. In its conclusion, it reaffirms the human being as the sovereign agent in final decisions. In the club, that means the machine helps, but a person decides and answers for the result.

On May 1, 2025, the church's South American headquarters launched 7chat.ai, “the Adventist AI,” as a beta version trained on official church content. The announcement itself admits that the assistant may offer inaccurate information and recommends checking answers against official sources and consulting pastors or church leaders on doctrinal questions. On October 6, 2026, the 7chat.ai address redirected to ia.adventistas.org.

Technology is powerful, but it can't replace the human touch.Pastor Jorge Rampogna, communication director for the Adventist Church in South America, Adventist News Network, 2025

📏Can Pathfinders use AI for Class and Honor requirements?

As of October 6, 2026, we found no South American Division rule specifically about artificial intelligence in completing requirements. The 2023 document leaves those details to each area: it says it does not intend to specify criteria for every form of use and asks church entities to create their own guidelines. The 2020 edition of the Pathfinder Club Administrative Manual does not mention the subject either.

The Manual, however, already has two rules that cover the problem. In section 4.1.4, on Class evaluation, it asks the Instructor to check whether the text was really written by the Pathfinder and not just copied and pasted from a website or a book, and it allows asking for an oral report. In section 4.2.2, on Honors, it states that what must NEVER happen is the use of pre-answered Honors, especially those found online.

In our reading, which carries no official weight, an AI-generated answer handed in as your own works like a pre-answered Honor: the tool changes, but the shortcut stays the same. The guides on pre-answered Honors and on the answered Class in PDF explain why a ready-made answer key does not pass the evaluation.

At the start of a Class or Honor, ask your Instructor whether AI is allowed and at what stage. Allowing it only for spell-checking or asking for everything by hand are both legitimate choices until the SAD publishes its own guidance.

✅How to use AI ethically: what's okay and what isn't?

One question settles almost every case: after using AI, were you the one who learned? If the tool explained and you understood, you studied. If it wrote and you just copied, the requirement has no real owner.

Behind that test is the Pathfinder Law. The Administrative Manual explains “keep a clear conscience” (the item known in English as “Keep a level eye”) as a commitment not to lie or deceive, and “Do my honest part” as duty done honestly. See also the guide on honesty and integrity.

CERT.br suggests a practice that matches the transparency the church voted for: if you used AI, say where and for what purpose. One line at the end of the work is enough, such as “I used AI to generate review questions and checked the answers in the Class book.”

Guidance based on the Administrative Manual (SAD, 2020) and CERT.br (2026). Not an official rule: check with your Instructor.

Read alsoGot scammed or had your account stolen?→

📖How can you use AI for Bible study?

With your Bible open beside you. AI can suggest the background of a book, questions about a chapter or passages on the same theme, but every reference has to be checked against the biblical text, because the tool can swap verses and blend quotations.

The Bereans are the model. Acts 17:11 (NIV) says they received the message of Paul and Silas “with great eagerness and examined the Scriptures every day to see if what Paul said was true.” And 1 Thessalonians 5:21 (NIV) sums up the attitude: “test them all; hold on to what is good.”

On doctrinal questions, the advice from the 7chat.ai announcement works for any tool: check official church sources and talk to your pastor or a leader. To build your own method, see the guide how to study the Bible at ages 10-15.

In devotions, “Keep the morning watch” opens the Pathfinder Law, and its value lies in your own time with God. AI can suggest a structure, but the reading and the prayer have to be yours.

🔒What should you never share with AI?

CERT.br warns that whatever you type into a chatbot usually goes to the company behind the AI and may be used to train the model, with a risk of showing up in answers to other people. Assume that anything you give the AI is public: if you would not put that information on a poster in the street, keep it out of the prompt.

In the club, that means never pasting into a prompt a full name, address, phone number, ID number (such as Brazil's CPF), password, face photo or medical form, yours or another Pathfinder's. To organize a list, replace names with generic labels such as “Pathfinder 1.” For leaders, medical forms, permission slips and photos of Pathfinders are data about children and teenagers; the rules for protecting them online are in the guide on Brazil's ECA Digital and what it changes for families, a Brazilian law in force since March 17, 2026.

Adjust the tool as well: CERT.br recommends turning off the use of your chats to improve the model, switching off activity history and using temporary chat when it is available. It also warns against turning AI into a therapist, because conversations with AI carry no professional confidentiality. If something is weighing on you, talk to your parents or your Counselor.

🎂What is the minimum age for ChatGPT and other AI tools?

Each service sets its own minimum age in its terms of use. The Terms of Use of OpenAI, the company behind ChatGPT, in force since January 1, 2026, require users to be at least 13 or the minimum age required in their country, and anyone under 18 needs permission from a parent or legal guardian. We cite this case as an example, not as a tool recommendation.

For schools in Brazil, the National Education Council (CNE), the body that sets rules for Brazilian education, approved on September 1, 2026 rules for AI use in the country's schools and universities and banned independent use by students up to 5th grade, with activities mediated by education professionals. The text then went to Brazil's Ministry of Education (MEC) for ratification.

For the club, the math is simple: Pathfinders aged 10 to 12 are below the 13-year minimum in that example, and older ones depend on their parents' permission. An AI activity during a meeting is safer when an adult leads it on the leader's device, without asking a Pathfinder to create an account.

Sources: OpenAI Terms of Use (Brazilian Portuguese version), UNESCO (2023) and Jornal do Comércio, September 1, 2026. Status checked on October 6, 2026.

🎭How can you tell if an image, audio clip or video is AI-generated?

CERT.br warns that scammers use AI to create deepfakes, posing as someone familiar or famous to ask for money, run misleading ads or spread false information. An audio message in the Unit group chat, in the Director's voice, asking for an urgent PIX transfer (Brazil's instant payment system) is exactly that kind of scam.

Before believing or sharing, ask the booklet's questions: whether the person would really say that, whether money, an advantage or personal data is involved, and whether there is urgency, a threat or an emotional appeal. Confirm through another channel. When in doubt, CERT.br says, assume it is fake. Proverbs 14:15 (NIV) already warned: “The simple believe anything, but the prudent give thought to their steps.”

In the Unit, the Counselor can set up a game with real news and fake edits, without using the group's faces. Those who want to go further will find at AdventSource, in the United States, an Artificial Intelligence honor at Skill Level 2, created in 2014; check with your regional coordinator whether it appears in the South American Division Honors Manual. In the South American catalog, the topic appears in the family of Computers and Internet Honors: Computers V asks for a report on how computing has evolved in areas such as artificial intelligence.