The Connectivity Paradox Holding Back Enterprise Agentic AI

AI agents have become the primary driver of enterprise productivity. According to a recent report by Salesforce, organizations currently use an average of 12 agents, with that number expected to climb 67% within the next two years.  Currently, 83% of organizations report that most or all teams and functions have adopted AI agents. However, business…

Read More

Single prompt breaks AI safety in 15 major language models

A single benign-sounding prompt can systematically strip safety guardrails from major language and image models, raising fresh questions about the durability of AI alignment when models are customized for enterprise use, according to Microsoft research. The technique, dubbed GRP-Obliteration, weaponizes a common AI training method called Group Relative Policy Optimization, normally used to make models…

Read More

The Data Gravity Problem Is Back, and AI Made It Worse

Data gravity never actually disappeared, so it may not be fully accurate to say it’s back. However, it definitely stayed quiet. Traditional data analytics workloads were forgiving enough that this problem was not catastrophic. Dashboards loading slightly slower, or reports running overnight were not a disaster. The system continued to work, even if it was…

Read More

OpenScholar Shows Why Grounded AI Matters for Scientific Research

Researchers from the Allen Institute for AI (Ai2) and the University of Washington have developed a new open-source AI model named OpenScholar that they claim can synthesize scientific literature and verifiable citations at a level comparable to a human expert. With millions of scientific papers published every year, it’s challenging to keep up with the…

Read More