Know The Hidden Risks of Public AI Models For Tribal Nations

As engagement with artificial intelligence accelerates, tribes nationwide have made AI awareness a priority.

Mark Hall, IT Security Manager for Potawatomi Casino Hotel, said AI awareness is critical for tribes, especially as technology evolves and many companies drive model development. 

"AI models often ingest inputted data to train future iterations," Hall said. "This practice risks exposing sensitive cultural knowledge, oral histories, or internal tribal governance records to the public domain."

Beyond data privacy, off-the-shelf AI tools often use datasets embedded with historical biases, which can lead to inaccurate automated decisions in tribal healthcare, housing allocation, or judicial systems, according to Hall. 

Hall said legal and regulatory friction presents another major challenge, as mainstream AI vendor contracts regularly dictate that legal disputes be settled in federal or state courts.

"This language can inadvertently bypass or undermine tribal court jurisdiction," Hall said, adding that general AI models do not natively understand specific treaty rights, federal Indian law, or unique tribal constitutions, leading to flawed or legally non-compliant automated outputs.

AI tools are also susceptible to hallucinations, with many models often generating responses based on perceived user expectations rather than verified facts. 

Beyond operational accuracy, AI is fundamentally altering business operations, echoing the shift the internet brought. 

Since Claude created Mythos, Hall said the volume of identified vulnerabilities skyrocketed in 2026. 

He said the era of managed monthly patching has passed, forcing organizations to scramble toward Continuous Patching Methods (CPM) to keep pace with modern threat vectors.

Hall shared strategies tribal leadership can use to stay prepared and proactive, including establishing governance by creating a tribal AI ethics board and updating vendor procurement policies to screen AI code algorithms.

Hall explained that the vendor vetting process can be one of the most vital steps.

"Demand full transparency on where data is stored, require vendors to sign strict non-disclosure agreements, and ensure all contracts explicitly respect tribal court jurisdictions," Hall advised.

Hall also said continuing education for tribal technology leaders on new model developments is key.

 

“Enroll leadership in Native-led tech forums and attend specialized federal Indian law and technology seminars,” Hall said. “Host internal workshops on basic AI risks and train staff to never input personally identifiable information (PII) into public AI tools and educate IT teams on auditing automated vendor software.”