Agentic AI's potential to reshape global business services is being met with a reality check, as deployment lags behind the hype, according to a recent VentureBeat report. While the technology, which allows AI to take goal-driven action, was expected to make significant strides in 2025, the fundamentals needed for widespread adoption are still missing, according to VentureBeat Contributing Editor Taryn Plumb.
The challenges of scaling agentic AI are evident, even as other AI applications make headway. Mastercard's fraud protection platform, Decision Intelligence Pro (DI Pro), is a prime example of AI's capabilities. The platform, which processes approximately 160 billion transactions annually, can identify suspicious transactions in milliseconds, according to Johan Gerber of Mastercard. This is crucial, as the network experiences surges of 70,000 transactions per second during peak periods.
Meanwhile, the rapid rise and fall of Moltbook, a Reddit clone for AI bots, highlights the current state of AI development. Launched on January 28, Moltbook quickly went viral, positioning itself as a social network where AI agents could interact. The platform, powered by the open-source LLM OpenClaw, allowed bots to "share, discuss, and upvote," according to its tagline. However, the project's ultimate impact remains uncertain.
MIT Technology Review's new AI newsletter, "Making AI Work," aims to provide practical guidance on AI implementation across various sectors. The newsletter will feature weekly case studies exploring tools and tips for generative AI deployment, addressing questions about how AI is being used in fields like healthcare, climate tech, education, and finance.
In other news, GitHub users can subscribe to updates and receive notifications about incidents via email or text message.
Discussion
AI Experts & Community
Be the first to comment