Dr. Kollin Napier, chief artificial intelligence officer for the Mississippi Artificial Intelligence Network (MAIN), told WXXV News 25 that the public should take artificial intelligence safety concerns seriously without treating an extreme-risk estimate as a prediction about the tools people use today. The interview with WXXV’s Megan Fayard, published September 11, focused on how testing, cybersecurity, governance and public understanding can support responsible progress.
Extreme warnings require context
Fayard asked Napier about a researcher’s estimate that advanced AI could pose an existential risk within the next decade. Napier distinguished that personal estimate about an extreme future scenario from a forecast about current systems.
“That number, that percentage, is one researcher’s personal estimate about an extreme future scenario,” Napier said. “It is not a prediction that this is going to happen, and it is not saying that AI people are using today carries that same level of risk.”
Napier compared AI development with road safety. Driving carries risk, so manufacturers improve vehicles, governments set rules and drivers learn how to use the technology. He said AI calls for the same continuing work on testing, safeguards and public understanding.
Comments Napier shared after the interview sharpened that position: “Fear is a poor strategy. Blind acceleration is, too.”
He argued that technical capability is only one measure of progress. Developers and users must also consider whether a system serves people, earns trust and leaves human beings accountable for consequential decisions.
Responsible adoption begins before deployment
Napier placed responsibility at two levels. Developers of frontier models must test and secure their systems. Businesses, public agencies, educational institutions and other users must govern how those systems interact with people, data and existing operations.
For organizations, responsible adoption includes:
- Knowing what data enters an AI system and where that data goes
- Limiting system access and autonomy according to the risk of the task
- Testing systems before deployment and evaluating them during use
- Monitoring for misuse, security failures and unintended behavior
- Keeping people accountable for high-impact decisions
The controls should match the use case. A tool that helps draft routine material poses different risks from a system that handles sensitive data, acts across connected services or influences a decision affecting a person’s rights, safety or access to essential services.
Before deployment, an organization should know what the system can access, which actions it may take, how its behavior will be evaluated and who is responsible for the result. Higher-risk uses require tighter limits and more demanding review.
Public understanding is part of AI safety
Napier also emphasized that people do not need technical credentials to begin evaluating AI. Basic literacy helps users recognize limitations, protect sensitive information, ask better questions and decide when an AI-generated answer requires verification or qualified human review.
Napier summarized the approach this way: “You do not have to blindly trust AI, and you do not have to be afraid of it. You should understand it.”
MAIN supports that work with no-cost, online AI courses, practical AI resources and policy and governance templates for Mississippi residents, educators, employers and public institutions.
Those resources give people a place to learn how AI works and give organizations a starting point for testing systems, protecting data, setting limits and assigning responsibility.
Watch the WXXV News 25 interview
This MAIN article summarizes and contextualizes the WXXV News 25 interview involving Napier and MAIN. Watch “AI’s rapid rise: Should humanity be concerned?”, reported by Megan Fayard and published by WXXV News 25 on September 11, 2026.
About MAIN: The Mississippi Artificial Intelligence Network is Mississippi’s statewide AI initiative, anchored at Mississippi Gulf Coast Community College. MAIN connects education, workforce, government, industry and community partners while providing AI learning opportunities, practical resources, responsible-adoption guidance and implementation support. Learn more about MAIN.