Why AI Trust Matters More Than Benchmarks (w/ Credo AI CEO)

Описание: In this episode of The Neuron Podcast, Credo AI founder and CEO Navrina Singh breaks down what AI governance really means and why benchmark scores alone won't save your company.

We explore Model Trust Scores, the White House's new AI Action Plan, open source safety concerns, and the challenge of governing autonomous agents. Plus, we test OpenAI's brand new open source model GPT-OSS live to see how it handles quantum physics, Shakespeare, and... cooking?

Topics covered:
• What AI governance actually means for businesses
• Why DeepSeek failed security requirements despite high benchmarks
• The White House AI Action Plan as a "moonshot moment"
• Open source AI from China - innovation vs security risks
• Governing AI agents that can access your financial accounts
• EU AI Act vs US deregulation approach
• Live demo of OpenAI's new open source model

Learn more about AI governance: https://credo.ai/resources

Subscribe to The Neuron newsletter: https://theneuron.ai

Special thanks to today's sponsor, Metaview: https://www.metaview.ai/

➤ CHAPTERS
0:00 - Intro
2:08 - What is AI governance?
6:39 - A message from today's sponsor, Metaview
7:44 - Model Trust Scores explained
11:09 - White House AI Action Plan
13:42 - Deregulation VS Governance
18:54 - Practical governance examples
26:51 - Open source models from China
30:26 - Agent-to-agent communication risks
40:18 - EU vs US Regulatory Approaches
45:51 - Will we see an AI lawsuit in 2025?
48:40 - Credo.ai and Credo.ai/resources
54:02 - Testing OpenAI's open source model
01:08:24 - Round Robin: Worrisome AI Unknown unknowns

Hosted by: Corey Noles and Grant Harvey,
Guest: Navrina Singh (CEO, Credo AI)
Published by: Manique Santos
Edited by: Kush Felisilda

Links to learn more:

Check out the Industry’s first use-case-based AI model leaderboard, Model Trust Score, enabling enterprises to select AI models based on several requirements: https://www.credo.ai/news/which-model...

Our article on Credo AI's new Model Trust Scores: https://www.theneuron.ai/explainer-ar...

The US AI Action Plan, for reference: https://www.whitehouse.gov/wp-content...

Navrina's blog on the AI Action Plan: https://www.credo.ai/blog/americas-ai...

Credo AI has just launched Credo AI Advisory Services, a strategic enablement offering built to make AI oversight actionable, measurable, and scalable for enterprises and government agencies alike: https://www.credo.ai/blog/credo-ai-la...

AI leaders Microsoft, IBM, Databricks, McKinsey, and Booz Allen Hamilton join forces with Credo AI to accelerate trusted enterprise AI adoption while ensuring compliance and risk management: https://www.credo.ai/blog/credo-ai-la...

Check out Credo AI at: https://www.credo.ai/