An AI chatbot is software that holds conversations with people and answers their input automatically. It uses artificial intelligence to understand questions in natural language and to produce responses. ChatGPT, Claude and Gemini are well-known examples.
Rule-based or AI-based?
Google Cloud distinguishes two kinds. Non-AI chatbots use scripted dialog with pre-programmed responses, such as an FAQ with fixed answers. AI-based chatbots use machine learning and large language models and cope better with varied wording. The term “AI chatbot” refers to the second group.
Shopify describes the usual sequence. The user asks a question, natural language processing identifies the intent, the system retrieves information, a language model writes the response, and the bot hands over to a human when needed.
Chatbots and SEO
Many AI chatbots fetch web pages to answer questions. Some providers use dedicated bots for this, which appear in server logs. ChatGPT-User fetches pages when a ChatGPT user asks for them, and OpenAI states that robots.txt rules may not apply to it because a user triggers the visit. Perplexity says the same for Perplexity-User, which visits a page to help answer a question and generally ignores robots.txt.
What a chatbot says about a brand therefore also depends on which pages it can fetch. Treat that as an interpretation. The bot behavior itself is documented by the providers.
Labeling under the EU AI Act
Article 50(1) of Regulation (EU) 2024/1689 requires providers to design AI systems that interact directly with people so that those people are informed they are dealing with an AI system. This does not apply where it is obvious to a reasonably well-informed, observant and circumspect person, given the circumstances and the context of use. The transparency rules apply from August 2, 2026. The Digital Omnibus, Regulation (EU) 2026/1744, did not postpone this date for Article 50(1).
This is not legal advice.
Example
An online shop embeds a chat that answers questions about delivery times. The answers come from a language model. At the start, the chat says that it is an AI assistant.
FAQ
Can an AI chatbot be wrong? Yes. Generative AI can hallucinate, meaning it may pass on wrong or made-up information, so important facts need to be checked.
More on the searchVIU blog
Sources
- AI chatbot – Google Cloudcloud.google.com
- AI chatbots for e-commerce – Shopifyshopify.com
- OpenAI crawlers and user agents – OpenAI Docsdevelopers.openai.com
- Perplexity bots – Perplexity Docsdocs.perplexity.ai
- Regulation (EU) 2024/1689 (AI Act), Art. 50(1) – EUR-Lex, English version, checked 7.10.2026eur-lex.europa.eu
- Regulation (EU) 2026/1744 (Digital Omnibus on AI), Art. 1 points (39) and (40) (Art. 111(4), Art. 113) – EUR-Lex, English version, checked 7.10.2026eur-lex.europa.eu
- How to use chatbots and what to know – Yahoo Tech (secondary)tech.yahoo.com
Last updated: