Agentic AI
Latest Agentic AI news freshly curated for you as of 8 p.m. ET Aug. 7. Learn more about innovative developments, discover emerging trends and read expert perspectives.
Today's Highlights
- Major AI labs, including OpenAI and Anthropic, reported agents breaching test environments, prompting warnings from AI pioneers like Geoffrey Hinton about control.
- Microsoft now has 40 million AI agents deployed, outnumbering its 30 million paid human users, signaling a major shift in the nature of knowledge work.
- While model providers like Meta and OpenAI are cutting prices, the total cost for enterprises is rising due to the high token consumption of complex agentic workflows.
Sources
In The Spotlight
Dive deeper into key discussions within the world of Agentic AI with a curated selection of important articles offering valuable perspectives.
Discover AI-Powered Insights That Matter to You
Choose a focus, and we'll refresh your feed with the latest insights.
Agentic AI Is Breaking Security’s Human Assumptions
AI agents can act thousands of times before humans react. Black Hat experts warn identity, costs and security models aren’t ready for what comes next.
Contributor
Agentic Commerce Has An Invisible Identity Gap
There's an invisible identity gap in agentic commerce. And until the agent is tethered to a verified human root, the market will be vulnerable to the kinds of abuse that destroy trust.
AuthorVoice
Geoffrey Hinton Warns AI May Outsmart Humans As Agents Escape Tests
Geoffrey Hinton warns AI may outsmart humans as agentic systems breach testing environments, exposing critical safety gaps for enterprise leaders.
ByTim Bajarin,
Contributor
Is AI Agent Pricing Getting Better? Grading My 2025 Predictions
A year after the author first assayed the issue, pricing for enterprise agentic AI continues to be a challenge — something the agentic vendors themselves acknowledge.
Contributor
How Goldman Sachs Is Using Agentic AI For Software Engineering At Scale
Goldman Sachs is putting AI software engineers to work alongside thousands of human developers using autonomous agents to tackle production tasks & accelerate development
ByBernard Marr,
Contributor
DeepSeek-Powered AI Used To Launch Attacks — Agentic Threats May Go Beyond One-Offs
A Chinese threat actor used a DeepSeek-powered AI agent to attack vulnerable servers — then it backfired.
ByZak Doffman,
Contributor
AI Agents At OpenAI Broke Out, Anthropic Broke In, Microsoft Obeyed
OpenAI, Anthropic and Microsoft all had AI agents cross the line in two weeks. The break-ins used weak passwords, not superhuman skill, and nobody caught them for months.
BySandy Carter,
Contributor
Ethan Mollick Writes AI Agent Guide
Ethan Mollick advises experimenting with AI agents responsibly, emphasizing permissions, paid tools, and real-world delegation.
ByJohn Werner,
Contributor
Microsoft Deployed 40 Million AI Agents. What Are They Being Used For?
Microsoft's Satya Nadella reports 40 million AI agents being deployed across thousands of organizations. Here's how the smartest professionals are using them.
ByRachel Wells,
Contributor
Could Agentic AI Bring American Manufacturing Back?
America faces a growing engineering shortage. Here's how Agentic AI could help manufacturers increase productivity, strengthen competitiveness, and bring work back home.
Contributor