Artificial intelligence (AI) is raising concerns in the field of archaeology, particularly regarding the misrepresentation of Indigenous knowledge and the spread of false historical narratives. A notable example is a photograph of large carved stone heads in a red desert, which appears near the top of Google Images when searching for "cultural heritage Australia." This image, however, is a fabrication created by generative AI (GenAI), a technology used in chatbots and image generators. While GenAI can expand access to knowledge, it also risks promoting inaccurate histories and stripping Indigenous Cultural and Intellectual Property (ICIP) of its context. Researchers are increasingly alarmed about how these systems often scrape and reuse Indigenous knowledge without proper consent or understanding. Pseudoscience in archaeology is not new, and it often relies on unproven theories or fringe narratives that lack scientific rigor. Concepts such as the "lost continents" of Mu and Lemuria, which have no basis in science, have long been used to justify harmful ideologies like eugenics and Social Darwinism. Pseudoarchaeology, which often draws on these same outdated ideas, tends to prioritize sensational stories over evidence-based research. These narratives can be more appealing to the public due to their sense of mystery and hidden truths, often overshadowing the slower, more meticulous work of real archaeology. Generative AI systems, which learn from the vast content on the internet, can inadvertently reproduce misinformation and fringe theories without discerning their validity. Unlike human researchers, AI does not assess evidence or verify claims. Instead, it predicts the next plausible word or image based on patterns in its training data, which can lead to the confident reproduction of falsehoods. This is compounded by the fact that Indigenous knowledge is deeply rooted in specific places, communities, and relationships with the land, making it fundamentally different from the statistical patterns that AI systems rely on. Indigenous knowledge systems are not just stories or cultural expressions; they include detailed environmental records, landscape knowledge, and data developed over generations through lived experience. These knowledges are relational and governed by specific protocols, rather than being passive information to be extracted. In response to these challenges, researchers in Australia and the Pacific have been working closely with Indigenous communities to build trust and use AI for purposes like language revitalization and cultural research. To address the issues raised by AI, the research community must ensure that Indigenous communities retain control over how their knowledge is used, through true collaboration rather than superficial consultation. This includes integrating Indigenous Cultural and Intellectual Property into research practices, ethics reviews, and funding decisions.