AI is being rolled out at scale across industries - but the safeguards haven't kept up. Bias in AI has improved less than 5% despite years of awareness, agentic AI is amplifying risks faster than companies can manage them, and most organizations still treat responsible AI as a compliance checkbox rather than a strategic advantage.
In this episode of AI Made Simple: The Transformation Series, I'm joined by Toju Duke - AI advisor, author, and former Google Responsible AI lead - who spent a decade at Google before founding Diverse AI, an organization building a 5-million-image dataset to represent communities invisible to current AI systems.
We discuss:
Why AI bias hasn't meaningfully improved and what's blocking progress
The security and accountability risks of deploying AI agents without safeguards
How underrepresented data creates real business risk, from recruitment failures to reputational damage
Why responsible AI should be part of your AI strategy, not a separate initiative
Podden och tillhörande omslagsbild på den här sidan tillhör
Valeriya Pilkevich. Innehållet i podden är skapat av Valeriya Pilkevich och inte av,
eller tillsammans med, Poddtoppen.
