AI

Marc Benioff says Salesforce will spend $300 million on Anthropic tokens this year, nearly all on coding. Here's what that signals about enterprise AI moving from experiment to operating cost.

OpenAI launched personal finance tools for ChatGPT Pro users, letting them connect bank accounts via Plaid. Here's what it actually does, what it can't do, and the privacy questions worth thinking through.

Microsoft's MDASH runs more than 100 AI agents in parallel to scan Windows code. In May 2026 it found 16 real CVEs, including 4 Critical RCEs, and scored 88.45% on the CyberGym security benchmark.

xAI launched Grok Build on May 14, a terminal-native coding agent with an 8-parallel-agent architecture and 2 million token context window. Here's what developers need to know.

Sinch surveyed 2,527 decision makers across 10 countries and found 74% of enterprises already rolled back deployed AI agents. The cause isn't model quality: it's the infrastructure layer most deployment plans skip.

Cerebras priced its IPO at $185 and closed up 68%, raising $5.5 billion in the year's largest tech debut. Its AI chip is built from a single silicon wafer, with OpenAI holding an 11% stake.

Cisco's Q3 earnings revealed $5.3 billion in AI infrastructure orders year-to-date and a forecast raised to $9 billion. Networking orders surged 50% and data center switching climbed 40%.

Apple is designing a new compliance framework that would allow AI agents to operate inside App Store boundaries, with details expected at WWDC on June 8 alongside a major Siri overhaul.

Real clips from Google's unreleased Gemini Omni video model are circulating publicly, showing in-chat video editing and strong prompt adherence just days before Google I/O 2026 opens on May 19.

Amazon retired Rufus and launched Alexa for Shopping, an AI assistant that monitors prices, schedules purchases, and buys things on your behalf, even on websites Amazon doesn't own.

Anthropic is in advanced talks to acquire Stainless for $300M+. Stainless builds the official SDKs for OpenAI, Google, Meta, and Anthropic itself. This would be Anthropic's first acquisition ever.

NVIDIA's open-source OpenShell runtime is now embedded into SAP's Business AI Platform as the security foundation for 200+ enterprise agents announced at SAP Sapphire.

AWS became the first cloud provider to offer Anthropic's native Claude Platform, with IAM auth, CloudTrail logging, and unified billing. Here's what's included and how it compares to Bedrock.

A GitHub repo with 137,000 stars exposed the system prompts behind Cursor, Windsurf, Claude Code, and 27 other AI coding tools. Here's what the hidden instructions actually reveal.

Microsoft tested 19 AI models on complex document editing across 52 professional fields. Frontier models corrupted 25 percent of content during long sessions. Adding agentic tools made outcomes worse, not better.

Google Chrome silently downloaded a 4GB Gemini Nano AI model to user devices without consent. Researcher Alexander Hanff flagged EU law violations and estimated carbon costs at 60,000 metric tonnes.

New data from 23,000 Kubernetes clusters shows enterprise GPU utilization averages just 5%. As AWS raises H200 prices by 15%, the cost of AI compute waste is becoming impossible to ignore.

Anthropic introduced dreaming to Claude Managed Agents on May 6, alongside outcomes grading and multiagent orchestration. Legal AI company Harvey saw task completion rates jump roughly 6x in early tests.

88% of enterprises reported AI agent security incidents last year. The cause: agents select tools from shared registries using descriptions no one verifies, and attackers found that gap before most security teams did.

ByteDance raised its 2026 AI capex 25% to $30 billion. Much of the increase goes toward domestic chips, including $5.7 billion in Huawei Ascend hardware, as US export controls reshape China's AI infrastructure strategy.

On May 7, 2026, EU lawmakers agreed on the Digital Omnibus on AI, amending the AI Act with timeline extensions, new deepfake prohibitions, and simplified compliance rules. Here's what changed.

OpenAI released three new voice models this week: GPT-Realtime-2, GPT-Realtime-Translate, and GPT-Realtime-Whisper. Here's what's technically new, how pricing works, and what developers can build with them.

NVIDIA backed IREN with $2.1 billion in potential investment to build 5 gigawatts of AI infrastructure. Here's what the deal structure means and why Sweetwater Texas is at the center of it.

Meta revealed plans for 'Hatch,' a personal AI agent, and a shopping tool for Instagram that completes purchases from Reels without leaving the app. Internal testing targets end of June; shopping tool aims for Q3 2026.

SAP, ServiceNow, Workday, HubSpot, and Datadog are all building access fees for AI agents. Per-seat billing is ending. Here's what the new model costs you and what to do about it.

Amazon's new AgentCore Payments lets autonomous AI agents make real micropayments mid-task using Coinbase and Stripe, built on the x402 HTTP protocol with full consent controls and spending limits.

Rivals turned partners: Anthropic just secured all compute at xAI's Colossus 1 supercomputer, adding 220,000 Nvidia GPUs to Claude's infrastructure while Claude Code rate limits double immediately.

ServiceNow and NVIDIA launched Project Arc at Knowledge 2026: an autonomous enterprise agent inside a sandboxed runtime, with every action logged in real time by AI Control Tower.

Anthropic has committed $200 billion to Google Cloud over five years, locking in 3.5 gigawatts of TPU capacity. It reveals just how fast Claude's demand has outgrown original plans.

Microsoft says Agent 365 is generally available, giving enterprise teams a unified control layer for AI agents across security, identity, and workflow systems.

At AI Ascent 2026, Andrej Karpathy framed a shift from one-shot prompt coding toward agentic engineering workflows, a change that could reshape how software teams ship and govern AI-assisted development.

The Pentagon’s May 1, 2026 move to bring eight AI vendors into classified network work is less about one contract win and more about a new procurement model that could reshape enterprise AI buying.

Mistral’s new remote agents in Vibe point to a larger shift in how coding assistants are used: less pair-programming at the keyboard, more parallel cloud execution with human review at decision points.

Cloudflare says its in-house inference engine now runs larger open models on fewer GPUs. That claim matters because inference cost and latency have become the main bottlenecks for many AI product teams.

Mozilla says AI-assisted analysis helped identify 271 Firefox vulnerabilities before release. The deeper signal for 2026 is how security teams must redesign triage, remediation, and governance workflows.

OpenAI’s latest infrastructure roadmap puts power, financing, and deployment speed at the center of the AI race, and it changes what enterprise teams should expect from cloud, model, and platform vendors this year.

Snowflake’s April 21, 2026 platform update puts coding assistance and data-grounded agents in one operating model, forcing enterprise AI teams to rethink ownership, rollout order, and cost controls.

OpenAI announced GPT-5.5 on April 23, 2026 and moved it into API availability on April 24, shifting how teams handle rollout sequencing, cost controls, and governance for multi-step AI workflows.

On May 1, 2026, CISA, NSA, and allied agencies warned that many AI agent deployments are over-privileged and under-monitored, urging tighter identity, access, and approval controls before scale.

Anthropic says 6% of sampled Claude conversations involve personal guidance requests, a behavior shift that forces product teams and enterprises to rethink AI trust, safety policy, and governance controls.

OpenAI says it has already secured 10 gigawatts of U.S. AI compute capacity, years ahead of its 2029 goal, signaling that power contracts and infrastructure access are now central to AI competition.

Lens says its new Lens Agents platform applies policy, identity, and audit controls across AI agents running in cloud stacks and employee desktop tools, signaling a governance-first shift for enterprise AI operations.

AWS's April 27 roundup puts Bedrock AgentCore CLI alongside major ecosystem signals, showing how enterprise AI agents are moving from pilots to repeatable delivery workflows.

OpenAI’s April 29 infrastructure update points to a bigger change than capacity growth, AI competition is becoming a power, data center, and financing race that will decide which teams can ship dependable products.

Anthropic and Amazon say they will build up to 5 gigawatts of AI compute capacity, a scale jump that could reshape model availability, pricing pressure, and procurement timelines for enterprise teams.

Stripe says merchants will soon be able to sell inside Google AI Mode and the Gemini app, a move that could shift AI shopping from demo behavior into measurable transaction flow.

NVIDIA says its new open Nemotron 3 Nano Omni model is designed to run multimodal AI agent workloads with lower inference cost, signaling a market shift from benchmark talk to deployment economics and operational fit.

Arm says new deployment and open-standards work announced at OCP EMEA 2026 aims to make AI agent infrastructure easier to run at enterprise scale.

CIS released three new AI security companion guides in April 2026, giving security teams concrete control mappings for LLMs, AI agents, and MCP-connected tools.

Cognizant's April 29, 2026 Astreya acquisition is a clear signal that major IT services firms are racing to own AI-first managed operations at enterprise scale.