AI safety and innovation can advance together. That was a central message from Dr. Chris Mattmann, Founder and President of Mattmann.AI, during his appearance with Zain Asher on CNN International’s One World with Zain Asher and Bianna Golodryga.
The conversation examined the growing scrutiny of advanced AI systems: when companies should hold back a release, how autonomous agents respond to human oversight, and whether the United States can compete with China while maintaining meaningful safeguards.
When AI Safety Changes a Release Decision
Asked about OpenAI’s decision to withhold GPT-6.1 Astra, Mattmann described the move as evidence that safety concerns can affect a company’s deployment decisions, even after substantial investment.
He explained the importance of alignment: work intended to guide AI systems toward human values, instructions, and behavioral boundaries. In the interview, he highlighted concerns about systems exceeding their authorization or becoming less responsive to human feedback.
For Mattmann, evaluating an advanced model involves more than measuring what it can accomplish. It also means examining how the system behaves when people correct it, limit its permissions, or review its actions.
Why Human Oversight Matters for Agentic AI
The discussion focused on agentic AI: systems that can use tools and take actions, rather than only generate responses. Mattmann emphasized that these capabilities make oversight and cybersecurity particularly important.
When an AI system can interact with networks and other computers, staying within its assigned task and authorization becomes a practical safety requirement. The interview explored why developers need to observe these systems closely and take concerning behavior seriously before releasing them more broadly.
AI Safety and Innovation in a Competitive World
Asher asked whether the United States could maintain its competitive position while building public trust and addressing AI risks. Mattmann’s answer was that both goals are achievable.
He pointed to international research collaboration and the continuing role of universities and research institutes in AI safety. In his assessment, this work continues even when governments differ in their regulatory priorities.
The conversation also addressed competition from Chinese AI developers, including lower-cost development approaches and model distillation. Mattmann argued that competitive pressure makes continued innovation important, while maintaining that safety deserves attention alongside that progress.
“The safety work is still happening with or without the government.”
— Dr. Chris Mattmann, CNN International
The Outlook for AI Oversight
The segment also examined industry warnings about advanced AI and the government’s response. Mattmann welcomed greater attention to safety and said he expects regulatory scrutiny to remain an important part of the conversation.
His position connected technological leadership with accountability: the United States can pursue AI development while taking human oversight, cybersecurity, and responsible deployment seriously.
Watch Chris Mattmann on CNN’s One World
Watch the interview with Zain Asher below, or view the clip directly on X.
Watch Chris Mattmann discuss AI safety and innovation with Zain Asher on CNN International’s One World.
Thank you to Zain Asher and the CNN International team for the opportunity to discuss these questions on One World.
Learn more about Mattmann.AI and our work connecting AI strategy, technology leadership, and practical implementation.