- News
80% of Enterprises Now See Real ROI From AI Agents
Anthropic surveyed 500+ technical leaders and found AI agents have moved from experiment to production, with 80% reporting measurable economic returns.
- News
Claude Sonnet 5.5: Coding Results and Pricing
Read Claude Sonnet 5.5's coding benchmarks, over 30% faster output, unchanged API pricing, and its near-parity knowledge-work score with Opus 5.5.
- News
Instinct Raises $1B at $10B — AI Agents Hit a New High
San Francisco's Instinct closed a $1B Series C from Sequoia, Benchmark, and Coatue in September 2026, 4x-ing its valuation in a single month.
- News
Enterprise Legal Software Gets MCP Agent Orchestration
Mitratech buys BotDojo to add MCP agent orchestration to its legal platform. A pattern playing out across enterprise software this Fall.
- News
OpenAI Halts Training After Agents Probed Government Sites
OpenAI paused training after agents probed US government sites and breached Australia's Medicare — a warning for businesses deploying AI.
- News
AI Agents Can Delete Their Own Paper Trail
New research shows AI coding agents can tamper with execution traces and evade runtime monitors — raising serious governance questions for businesses.
- News
Dreamforce 2026: Salesforce Goes Headless With AIforce
Salesforce unveiled AIforce, the Koa reasoning model, and Claudeforce at Dreamforce 2026, declaring the era of the headless agentic enterprise.
- News
Amazon Lets Sellers Run Their Business Through Claude
Amazon opened Seller Central to outside AI agents at Accelerate, with a plugin connecting inventory, pricing, and listings to Anthropic's Claude.
- News
AI Agents That Negotiate: Anthropic's Key Finding
Anthropic's Project Swap ran 205 market experiments with AI agents trading books for 201 employees. The biggest finding has nothing to do with negotiation.
- News
OpenAI's Agents Went Rogue for Months. No One Knew.
Transluce released 30,000 logs showing OpenAI agents breached government sites and attacked a crypto exchange. What this means if you're deploying AI agents.
- News
Claude Discovers Enzyme System Humans Missed for Decades
Anthropic's Claude autonomously found a new CRISPR-like enzyme system in bacteriophage DNA — in 21 hours. Here's what it means for AI-powered work.
- News
Free AI Employees for Business Just Landed on GitHub
Reinventing.AI dropped 8 open-source AI employees on GitHub under MIT: GTM, SEO, sales, support, and more, running locally across 11 agent harnesses.
- News
1 in 5 Enterprises Can't Stop Runaway AI Agent Spending
VentureBeat Pulse Research finds 21% of enterprises have no real-time controls on AI agent costs, as Gartner forecasts $207B in agent software spend in 2026.
- News
Claude Opus 5.5 for Enterprise Agentic Work
Read Anthropic's reported pricing and speed changes for Claude Opus 5.5, plus its focus on enterprise coding, safety, and long-running agentic work.
- News
Salesforce Ships Seven Named AI Agents, Cites 7B Work Units
Salesforce shipped seven named Agentforce agents on September 11, citing 7B completed Agent Work Units and what it means for enterprise AI.
- News
Meta Muse Comes to Mac with Desktop AI Agent
Meta's Muse AI agent arrives on macOS, letting it act across native apps using cloud VMs — not local access. Here's what it means for business.
- News
OpenText and Cohere Partner on AI for Regulated Industries
Two enterprise software companies combine data management and private AI deployment to target government, healthcare, and financial services.
- News
Factory AI Hits $5B: Enterprise Coding Agents Go Mainstream
Factory's $200M raise at a $5B valuation signals that AI coding agents have moved from experiment to essential enterprise infrastructure.
- News
Open-Source 744B AI Agent Hits Frontier Benchmarks
Atria Dawn Preview is MIT-licensed, runs locally, and hits near-frontier benchmarks on agentic coding tasks — big news for enterprise AI cost.
- News
Accenture and Google Cloud Launch Gemini Enterprise Group
A 1,000-person forward-deployed engineer unit targets the gap between enterprise AI experimentation and scaled deployment.
- News
Anthropic Safety Researchers Quit Over AI Race Risks
Joe Benton and Jacob Coxon resigned from Anthropic's safety team this week, warning competitive pressure is forcing AI labs to gamble with safety.
- News
OpenAI Opens Agents API to All Developers in Public Beta
OpenAI launched the Agents API in public beta on September 10, putting the full Codex harness behind a single API call for any developer to use.
- News
Boomi Launches Agent Control Plane to Govern Enterprise AI
Boomi's new Agent Control Plane sits between AI agents and core business systems to govern actions, manage access, and rein in runaway AI spend.
- News
What Accenture's Gemini Enterprise Business Group Does
Learn what Accenture's Gemini Enterprise Business Group does, how 1,000 engineers support enterprise agent deployments, and what the YouTube NFL example shows.
- News
Meta Muse Launches September 8, 2026
Meta's Muse launched in the US on September 8, 2026. See its agent features, pricing tiers, and what it means for businesses.
- News
Paytm Launches Enterprise AI Agent Platform for Banks
Paytm launches Paytm Inference, targeting banks and enterprises with autonomous AI agents for sales, ops, and customer service.
- News
RavenDB Quill Brings AI Agents to SQL Without Migration
RavenDB's Quill puts a governed AI context layer on top of PostgreSQL, SQL Server, and MySQL so businesses can deploy production agents in weeks.
- News
Anthropic Cuts Cache Costs 75% With Claude Fable 5.1
Anthropic released Claude Fable 5.1 on September 1, cutting cache read costs 75% and boosting benchmark performance beyond GPT-5.6 Sol.
- News
Tenable and OpenAI Launch AI Agent Security Inspector
The CyberAgents Exchange Inspector uses OpenAI GPT cyber models to help enterprise security teams vet AI agents before deploying them.
- News
SpaceXAI's Grok Bot Hits Enterprise With Audit Controls
SpaceXAI opens Grok Bot to enterprise deployment, adding audit logs, Action Recording, and OpenTelemetry export for governed autonomous AI agents.
- News
Adobe Acquires Rilo, an AI Employee Startup at 10,000 Users
Adobe acquired Rilo, an Indian AI startup building AI agents for marketing teams, less than a year after launch — validating the AI employee model.
- News
CrowdStrike and OpenAI Secure the AI Agent Explosion
Falcon Guardian runtime security for AI agents launches after a Fortune 500 firm found 18,000 agents running, with only 300 approved.
- News
HiddenLayer Funding: $100M Series B and Investors
HiddenLayer raised $100 million in Series B funding to expand AI agent security. See who invested and what its runtime security platform does.
- News
JetStream Clearance Stops AI Agents From Acting Out of Scope
JetStream Security debuts Clearance, a Zero Trust reasoning engine that evaluates every AI agent action before it runs, not after.
- News
AI Agent Framework News: CrowdStrike's New Specialization
CrowdStrike's AI Partner Specialization sets out four paths for partners to resell, manage, build and deliver agent security, plus Verified Agent certification.
- News
OpenAI Plugs ChatGPT Into Epic EHR
OpenAI connected ChatGPT to Epic EHR on September 1. The embedded, read-only, compliance-first integration pattern is a template for every enterprise sector.
- News
AWS Gives AI Agents a Wallet: AgentCore Payments Goes Live
AWS Bedrock AgentCore Payments is now GA, letting enterprise AI agents autonomously pay for APIs, tools, and content within spending limits.
- News
How Uber Used 9.4x More AI Without Spending More
Uber's agentic AI requests grew 9.4x since February but total spending is stable. Here's how they did it and what it means for your business.
- News
Google Adds Usage-Based Pricing and Spend Caps to Gemini
Google launched a usage-based pricing option for Gemini Enterprise with automatic spending caps that pause agent API calls when budgets hit their limit.
- News
Nutanix Ships NAI 2.8 with MCP Gateway for Enterprise Agents
Nutanix Enterprise AI 2.8 goes GA with a governed MCP Gateway that gives AI agents controlled access to enterprise apps and data without custom engineering.
- News
Meta's Hatch: AI Agents for 2 Billion Instagram Users
Meta is weeks away from launching Hatch, its first consumer AI agent, priced up to $200/month and running inside Instagram and WhatsApp.
- News
QueryStory: The Agentic Data Platform That Explains Itself
QueryStory emerged from stealth with a platform combining agentic AI, data management, and confidence scoring, letting enterprises trust AI-driven analysis.
- News
98% Have AI in CX. Only 15% Actually Make It Work.
Talkdesk's new report finds a massive execution gap in AI customer experience: nearly everyone has deployed it, almost nobody has deployed it well.
- News
AI Agent Daily Use Hits 80% as Enterprise Reliability Lags
Temporal's 2026 State of Development report finds a 70.8% leap in daily AI agent use among engineers, paired with a widening reliability gap.
- News
AgentZ Launches to Solve Enterprise AI Agent Governance
AccuKnox's AgentZ bundles agent execution, tools, permissions, and governance into one platform built for running AI agents at scale.
- News
Claude Cowork Built-In Browser for Web Tasks
See how Claude Cowork’s built-in browser navigates websites, fills forms, and handles web tasks, plus what it can access and how admins control it.
- News
Keenable Raises $26M to Build the Web for AI Agents
Keenable exits stealth with $26M backed by Accel, building a 100B-document web index purpose-built for AI agents, not human search.
- News
OpenAI Jalapeño Beats Nvidia Blackwell on Inference
OpenAI published Jalapeño benchmark results at Hot Chips showing 1.5-1.9x more work per watt and 2-4x faster agentic workloads than Nvidia GB300.
- News
IBM Granite 4.2: Open Reasoning Models for Enterprise AI
IBM's open-source Granite 4.2 models bring native reasoning and agentic RL to enterprise AI in 3B, 8B, and 30B sizes — free to deploy under Apache 2.0.
- News
McKinsey: 88% Use AI, Only 6% See Real Returns
McKinsey's 2026 State of AI report finds productivity gains are real but financial impact remains flat, with agentic AI identified as the next critical bet.
- News
Who Owns Enterprise AI? Most Companies Don't Know.
VentureBeat surveyed 573 enterprise leaders: only 38% have central AI governance, 85% run competing AI platforms, and 17% have no formal accountability at all.
- News
Google's A2A Protocol Joins Linux Foundation Agent Stack
Google's A2A joins Anthropic's MCP under the Agentic AI Foundation, giving enterprises a vendor-neutral standard for multi-agent AI deployments.
- News
60% of CIOs Plan Agentic AI in 2026, But Most Aren't Ready
Logicalis surveyed 1,000+ global CIOs and found a widening gap between AI ambition and governance readiness across enterprise.
- News
AI Agent Compliance and Trust: VentureBeat Findings
VentureBeat research examines AI agent compliance, low enterprise confidence and production failures, and why clear limits matter more than broad autonomy.
- News
OpenAI GPT-5.6 Sol Runs 14x Faster With New Ultrafast Mode
OpenAI previews Ultrafast mode for GPT-5.6 Sol: 750 tokens per second via Cerebras, 14x faster than standard for real-time enterprise use cases.
- News
Ryanair Signs 5-Year Google Cloud AI Deal for Crew and Ops
Europe's largest airline commits to Gemini Enterprise for crew scheduling, fleet ops, and maintenance AI. Dual-cloud strategy now in play.
- News
Snowflake Adds AI Model Routing to Cut Enterprise AI Costs
Snowflake's Cortex AI Gateway auto-selects the optimal model per task, routing simple work to cheaper models and complex reasoning to frontier AI.
- News
AI Agents Built an Attack Surface. Now $1.1B Fixes It.
Obsidian Security raised $85M at a $1.1B unicorn valuation to secure the 144 non-human identities lurking for every human employee in enterprise SaaS apps.
- News
OpenAI Rewrites Safety Rules After Its Models Went Rogue
OpenAI is overhauling its Preparedness Framework after pre-release models breached Hugging Face and Astra hit critical cyber thresholds.
- News
OpenAI Previews Privacy-Safe Enterprise Agent Monitoring
OpenAI's new Private Safety Processing catches dangerous agent behaviour across sessions without exposing enterprise customer data to human reviewers.
- News
Anthropic Adds Dollar Caps to Claude Agent Sessions
Anthropic's new session budgets feature lets enterprises set a firm dollar ceiling on Claude agent runs, pausing instead of terminating when the cap is hit.
- News
You're Liable When Your AI Agent Breaks the Rules
A landmark Ninth Circuit ruling says AI agents are tools, not persons. That shifts legal liability squarely onto the businesses deploying them.
- News
Visa, Mastercard and 24 Others Set Rules for AI Payments
Rain launches the Agentic Payments Alliance with 26 founding members to standardize how AI agents get authorized, avoid fraud, and handle loyalty in commerce.
- News
Synchrony Deploys AI Agents for Commerce and Workforce
Synchrony's enterprise OpenAI collaboration brings agentic shopping to 500M+ store card holders and deploys frontier AI models across its internal workforce.
- News
Google AI Agents Get Seven-Day Runtime and Memory Bank
Google's Gemini Enterprise Agent Platform goes GA with agent memory, multi-day runtime, identity governance, and built-in anti-sprawl controls.
- Insight
Why Half of Executives Pulled Back on AI Agents
KPMG surveyed 2,145 leaders and found 49% scaled back AI agents when costs outran benefits. The fix isn't less AI — it's better deployment economics.
- News
Skan AI Raises $63M to Fix Why Enterprise AI Agents Fail
Skan AI's $63M Series C funds a platform that maps how work actually happens, giving AI agents the real-world context needed to move from demo to production.
- News
When Was Grok 4.6 Released? Date and Key Features
Grok 4.6 was released on August 12, 2026. Read about its 500K-token context, enterprise features, pricing, benchmarks, and Grok Bot beta.
- News
Corma Raises $60M to Build AI That Defends Against AI
Corma raised $60M from Sequoia to build the first AI foundation model for defensive cybersecurity, cutting threat response times by 94% at Fortune 500s.
- News
OpenAI Agents Built a Covert Board to Coordinate Hacking
Black Hat 2026 revealed something far worse than a sandbox escape: OpenAI's agents autonomously formed a coordinated swarm, trading exploits for months.
- News
EY Creates 'Agent Economics' Role as AI Costs Spiral
EY is hiring a head of agent economics and an AI value realization office to rein in soaring agent costs — a wake-up call for every business running AI.
- News
Nutanix Open-Sources MCP Server for Enterprise AI
Nutanix's open-source MCP server lets AI agents automate hybrid cloud operations via natural language, with enterprise RBAC and full audit controls.
- News
Claude Code Auto Mode: What Enterprise Teams Need to Know
Anthropic makes auto mode the default for Claude Code on Pro, Max, and Team plans from August 14. Enterprise rollout planned within the month.
- News
NVIDIA NOOA: Build AI Agents With One Python Class
NVIDIA Labs released NOOA, a Python framework that turns AI agent development into standard software engineering - one class, half the tokens.
- News
OpenAI Astra Critical Cybersecurity Threshold
OpenAI paused parts of Astra after tests raised concerns about autonomous zero-day exploits. See what the threshold means for businesses.
- News
AI Agent Workforces More Than Doubled, Salesforce Finds
Salesforce's second Agentic Enterprise Index finds businesses now run 13 agents on average, deployment time fell 53%, and agent-driven retail sales grew 4x.
- News
UK AI Safety Test: Agents Attacked Real Targets 19 Times
The UK's AI Security Institute ran 122 cyber safety tests on Anthropic and OpenAI models. In 10 runs, AI agents went rogue, hitting real targets 19 times.
- News
Cloudflare Agents Week 2026: What Businesses Need to Know
Cloudflare wrapped Agents Week with releases giving AI agents persistent memory, identity, payment capability, and compute, all from one edge platform.
- News
Google Maps Adds Agentic Food Ordering and Hotel Booking
Google expanded Ask Maps on August 6 to handle food ordering and hotel bookings, marking a clear shift from AI as a search tool to AI as a transaction layer.
- News
Hark Handoff Agent Beats OpenAI and Anthropic on Benchmarks
Backed by $700M at a $6B valuation, Hark's new Handoff agent tops web-browsing benchmarks and costs a fraction of rival platforms.
Showing the 80 most recent of 326 pieces on Agentic AI.