keynews.ai

Key AI news stories in enterprise and tech.

For practitioners and enthusiasts

Full KeyRank · 100 stories · This Month

#StoryPillarKeyRankSource
1OpenAI’s rogue AI model incident was worse than we thoughtAgents82The Verge AI
2Gemini went rogue, hacked three companies, and Google hid itAgents78The Verge AI
3Google’s Gemini agents hacked three companies in new AI safety incidentAgents78Financial Times Technology
4Google's Gemini becomes latest AI model to break out and hack computer systemsAgents78CNBC Technology
5OpenAI agents attacked RubyGems back in MayAgents78Simon Willison
6OpenAI's new Agents API gives developers the infrastructure behind Codex and ChatGPTAgents78The Decoder
7OpenAI Launches the Agents API in Public Beta, Putting the Codex Harness Behind One API CallAgents78MarkTechPost
8Introducing the Agents APIAgents78OpenAI Blog
9Google says attackers used AI agents to steal credentials in under six hoursAgents78SiliconAngle
10Another swarm of OpenAI agents reached the open internet without the frontier lab’s knowledgeAgents78TechCrunch AI
11Anthropic Opens a Research Preview of the Model Hardware Standard (MHS): A Shared Specification for AI Agents to Safely Operate Physical DevicesAgents78MarkTechPost
12Google Deepmind's AI Co-Scientist now plans experiments, runs lab equipment, and writes scientific papersAgents78The Decoder
13Anthropic launches tool that can manipulate laboratory toolsAgents78Financial Times Technology
14Claude, Codex, and Hermes installed unowned code inside corporate networksAgents78Ars Technica
15Google’s Gemini is the latest AI model to hack other companiesAgents75TechCrunch AI
16Google's Gemini also accidentally hacked three real companies during security testingAgents75The Decoder
17Security researchers used Claude to help them hack into OpenAIAgents75The Verge AI
18Researchers link another hacking campaign to OpenAI agentsAgents75SiliconAngle
19AI governance moves from observability to provable controlAgents72SiliconAngle
20Google Deepmind's Dream-RSI helps AI agents improve by “dreaming” about past attemptsAgents72The Decoder
21Meta Launches Muse for Mac: A Personal AI Agent That Works Across Your Files, Mail, Messages, Calendar and NotesAgents72MarkTechPost
22Presentation: Context Engineering at LinkedIn: How We Built an Organizational Context Layer for AI Agents with MCPAgents72InfoQ AI/ML
23Cybersecurity researchers gain access to OpenAI’s GitHub repository using ClaudeAgents72SiliconAngle
24Google’s new ‘CC’ is an AI agent that helps families run their householdsAgents72TechCrunch AI
25DoorDash Uses Multi Agent LLMs to Clean up 60,000 Feature FlagsAgents72InfoQ AI/ML
26Meta’s Muse hits Mac, letting the AI take actions on your computerAgents72TechCrunch AI
27Anthropic keeps pushing Claude Code toward autonomous coding with new parallel agent workflowsAgents72The Decoder
28AI agents erase the paper trail, reshaping audit assuranceAgents72SiliconAngle
29Self-generated prompt injections in compaction summariesAgents72Simon Willison
30The fix for rogue AI agents could be more AIAgents72TechCrunch AI
31AI agent swarms are a massive waste of tokens with zero quality gain, says OpenAI Codex developerAgents72The Decoder
32Repeated VM Escapes By GPT-5.6-Cyber Based Agents Prove VMs and OS' Require Better MaintenanceAgents72InfoQ AI/ML
33OpenAI unveils new framework for reporting ‘AI misalignment’ as it reveals six more worrying incidentsAgents72SiliconAngle
34Stanford Researchers Release Paper2Agent: Turning Research Papers Into AI Agents That Reproduce Results and Run on New DataAgents72MarkTechPost
35OpenAI Creates a New Framework to Disclose Bad AI BehaviorAgents72Wired AI
36Your AI agents can now control your Google Home devicesAgents72TechCrunch AI
37Google will now let any AI agent run your smart homeAgents72The Verge AI
38Adecco Group announces Salesforce’s Agentforce Coworker rollout across 40 countriesAgents72Salesforce Newsroom
39Inside OpenAI’s agentic software factoryAgents72Pragmatic Engineer
40Salesforce announces AIforce, unlocking the power of its platform using composable agentsAgents72SiliconAngle
41StackGen launches Autonomous Operations Factory to govern production agentsAgents72SiliconAngle
42Exaforce adds a kill switch for AI agents that go rogueAgents72SiliconAngle
43Splunk rebuilds its platform around agentsAgents72SiliconAngle
44Agent-net Open Sources Webagent: A Go Harness That Turns Any Website into a Guarded AI AgentAgents72MarkTechPost
45Agent Harness vs Agent Framework vs MCP: Which Layer Owns the Loop, State, Tools, Permissions, and RecoveryAgents72MarkTechPost
46Akuity gives AI agents a governed path into productionAgents72SiliconAngle
47AI agents blew the whistle on their cheating colleaguesAgents72MIT Technology Review
48Why Most Enterprise Agent Pilots Never Reach DeploymentAgents72AI News
49Supio’s long-horizon agents point to a new operating model for law firmsAgents72SiliconAngle
50Context Engineering Inside the Harness: 4 Mechanisms That Beat Context Overflow and Goal Loss on Long-Horizon TasksAgents72MarkTechPost
51AWS Introduces Pizza Bot: An Open Source Inbox for Background AI AgentsAgents72MarkTechPost
52Perplexity trusts GPT-6 Astra with end-to-end systemsAgents72OpenAI Blog
53Anthropic spent this week in hot water over cybersecurityAgents72The Verge AI
54Salesforce introduces Enterprise AI Harness, AI Control PlaneAgents72SiliconAngle
55Extreme Networks’ Agent ONE Coworker moves AI networking from dashboards to answersAgents72SiliconAngle
56Amazon makes its agentic AI platform Quick generally available for desktop on Windows and macOSAgents72SiliconAngle
57Meta’s AI agent Muse is now the No. 2 app in the USAgents72TechCrunch AI
58Muse can shop, write emails, and negotiate prices for users, all through WhatsAppAgents72The Decoder
59General Robotics’ GRID platform engineers itself, cutting robot setup to hoursAgents72SiliconAngle
60Google Open-Sources Mantis: A Modular Skills Toolkit That Lets Coding Agents Find, Reproduce and Patch VulnerabilitiesAgents72MarkTechPost
61Viral AI assistant Instinct now has its own email addressAgents72TechCrunch AI
62Meta debuts its ‘secure by design’ personal AI agent MuseAgents72SiliconAngle
63Meta Introduces Muse, a Personal AI Agent That Runs on Its Own Dedicated Secure Cloud ComputerAgents72MarkTechPost
64How one hedge-fund manager built his firm to be powered entirely by AI agentsAgents72CNBC Technology
65Meta Releases Muse, a Personal AI Agent With Privacy ‘Built Into It’Agents72Wired AI
66Meta debuts its Muse AI agent. Will consumers trust it?Agents72TechCrunch AI
67How GPT-5.6 Sol helps run quantum computing experimentsAgents72OpenAI Blog
68Meta bets on AI agent Muse to catch up in AI raceAgents72The Verge AI
69Meta pushes into personal AI agents as company faces public reckoning over privacy and safetyAgents72CNBC Technology
70GitLab Warns That AI Agent Sandboxes Are Only as Secure as Their Network AccessAgents72InfoQ AI/ML
71OpenAI reports AI "research interns" and warns about its own pace at the same timeAgents72The Decoder
72Import AI 472: DeepMind’s cheating math agents; populist AI policies; and Forethought theorizes a nightwatchmanAgents72Import AI (Blog)
73OpenAI to set misalignment disclosure rules after agents took over a wikiAgents72SiliconAngle
74Google Mantis: An Agentic Vulnerability Scanning Harness for Reducing False PositivesAgents72InfoQ AI/ML
75Research acceleration: The view inside OpenAIAgents72OpenAI Blog
76UC Berkeley Researchers Release CUA-Lite, an Open Platform Unifying Sandboxes, Data, Evaluation and RL for Computer-Use AgentsAgents72MarkTechPost
77OpenAI confirms ‘wiki incident,’ says it’s ‘working on a framework’ for more disclosureAgents72TechCrunch AI
78Deepmind put 100 AI agents in a room and they sorted into cheaters, converts, and whistleblowersAgents72The Decoder
79OpenAI admits its disclosure practices need work after its autonomous agents hacked a German wikiAgents72The Decoder
80OpenAI's GPT-6 Astra hallucinates less but remains vulnerable to hidden prompt injectionsAgents72The Decoder
81Meta AI Released Muse Spark 1.3: An Agentic Coding Model That Uses ~20% Fewer Tool Calls and ~25% Fewer Tokens Than Muse Spark 1.2Agents72MarkTechPost
82Anthropic Released Claude Commerce Agents: An Apache-2.0 Blueprint for Shopping and Merchant Agents Across Retail, Travel, Telecom and EntertainmentAgents72MarkTechPost
83Hugging Face attack is a wake-up call about the risks of AIAgents72Financial Times Technology
84CrowdStrike builds an identity provider for AI agents, not humansAgents72SiliconAngle
85CrowdStrike’s Falcon Guardian shrinks an AI agent’s blast radiusAgents72SiliconAngle
86Cursor Cloud Agents can now run in Vercel SandboxAgents72Vercel Blog
87Capsule Security fine-tunes Nvidia Nemotron models to stop rogue AI agentsAgents72SiliconAngle
88Qwen Developers Open-Sources zg (zvec-grep): A Local-First Search Layer Unifying ripgrep, BM25, and Vector SearchAgents72MarkTechPost
89CrowdStrike gives AI agents an identity provider, parallel SOC investigations and package blockingAgents72SiliconAngle
90Perplexity Releases Hybrid Compute on Mac: Cloud Agents Orchestrate Down to a Local Model, Gated On DeviceAgents72MarkTechPost
91When agents move at machine speed, security teams lose their lag timeAgents72SiliconAngle
92AIR Security launches with $50M to build a firewall for AI agentsAgents72SiliconAngle
93Breakout time hits zero as CrowdStrike unveils autonomous red teaming: theCUBE’s Fal.Con 2026 day one keynote analysisAgents72SiliconAngle
94The packet path becomes the place to catch shadow AI before it spreadsAgents72SiliconAngle
95NVIDIA and CrowdStrike Strengthen Agentic Cybersecurity FrontierAgents72NVIDIA Blog
96Delegated authority turns the trusted AI agent into the security problemAgents72SiliconAngle
97How t54 built a trust layer with Amazon Bedrock AgentCore paymentsAgents72AWS Machine Learning Blog
98Researchers from Princeton, Ant Group and Stanford Introduce AQuA: A Two-Part Agentic Framework for Autonomous Factor Discovery and Model Development in Quantitative FinanceAgents72MarkTechPost
99AIR raises $50M to help companies vet the skills and add-ons AI agents useAgents72TechCrunch AI
100As AI agents take on enterprise tasks, companies face a new battle over access and controlAgents72SiliconAngle