Agentic reliability and evaluations : Enterprises that got burned by a bad eval are the most likely to remove humans from the loop, not the least

New VentureBeat Pulse data: trust in automated agent evaluation nearly tripled in July, but enterprises burned by a bad eval are removing humans fastest.

Agent context layers: Enterprises governing their AI data are catching twice as many bad answers as the ones who aren't

A VentureBeat survey of 101 enterprises found 68% traced a confident, wrong AI agent answer to bad context. Governed semantic ...

Agentic security: Enterprises enforce agent permissions two-thirds of the time — and isolate high-risk agents less than one in five

A VentureBeat survey of 116 enterprises finds most enforce AI agent permissions, but fewer than one in five isolate high-risk ...

Infrastructure and compute: Enterprises are buying AI compute for speed while flying blind on what it costs

New VentureBeat Pulse Research: 66% of enterprises run AI in production, but fewer than half can rigorously track what their AI compute actually costs.

Solo VC Ozi Amanat - K2 Global Raises $200 Million in Capital Commitments to Invest in Category-Defining AI Companies

K2 Global backs frontier technology companies including Neuralink, xAI, SpaceX, Shield AI, Tenstorrent, Synchron, Lambda, ...

SpaceXAI's Grok Bot turns agents into persistent digital coworkers that can operate your apps for $120-per-month

Grok Bot's differentiation is therefore less about proving that AI can operate software than packaging computer use, persistence, workflow learning and multi-agent coordination into something ...

Nvidia's Switchyard router reshuffles AI models mid-task, cutting task costs to a third in its own tests

Nvidia's Nemotron 3.5 Lightning pairs with its NeMo Switchyard router, which reassigns models mid-task and cuts task costs to a third in Nvidia's own tests.

LTX-2.5 can generate a 10-second AI video from an image in just 6.8 seconds on Nvidia superchips — and it's open weights

Because the weights are open, that entire experimentation phase can happen on a company's own hardware, with no ...

Mistral AI wants to build 1 gigawatt of European compute by 2030 — and lock in customers now.

SLA-backed regional inference, five-year European Compute Unit contracts, hosting China's GLM-5.2, and 1 gigawatt of compute ...

Why AI-driven purchase intent so rarely becomes a completed sale

When an AI assistant recommends a product or brand, it generates something valuable: a purchase-ready consumer with high ...

Your AI agent may be ready. Your sales motion probably isn’t.

Nearly every software company is pouring resources into agent development. Far fewer are rethinking the path from discovery ...

ZoomInfo Data Now Integrated With Microsoft Copilot Studio and Available in Microsoft 365 Copilot, Dynamics 365, Excel, and Word

ZoomInfo's headless GTM context layer now connects to Microsoft 365 Copilot, surfacing 500M contacts, 100M companies, and billions of signals inside D ...