Stay updated with the latest news about AI agents, autonomous AI, and automation tools.
Twilio's per-minute overage billing and OpenAI's spend-ceiling feature, built after $800 bug-driven bills, show how to price AI agent usage spikes.
A group of so-called “swarmchasers” who track droves of rogue AI agents say they have uncovered an incident involving Chinese AI. The researchers, who have discovered high-profile attacks involving OpenAI, found a group of agents trying to gain access to the Chinese maps website Amap while avoiding the site’s anti-bot restrictions. The behaviour echoes that of the OpenAI agents that escaped safety guardrails and attempted to access data from sites including those of the Australian government. ## Small but powerful, we’re taking on the fight for fairness every single day ### Your payment #### Card Details Rowan Howard-Jones, the group’s team leader, said the researchers analysed a dataset disclosed as part of the investigation into OpenAI agents’ now-notorious Hugging Face hack to find markers of agent swarms: “Then we found a bunch of Chinese agents.” ## Hear about our latest investigations, before they break It is the latest in a series of reports from independent researchers that have revealed previously unknown behaviour by groups of AI agents, ranging from relatively benign attempts to dodge access restrictions through to hacking government websites. Jack Cable, co-founder at the AI cybersecurity firm Corridor, discovered a series of attempts by OpenAI agents to access US, Canadian and Australian government websites. ### More From Big Tech [\ **OpenAI agents hijacked a website. Why didn’t the company tell anyone?** \ 11 September 2026](https://www.thebureauinvestigates.com/stories/2026-09-11/openai-agents-hijacked-a-website-why-didnt-the-company-tell-anyone)
## AI agents are here. How should designers feel?  The forthcoming Muse Charm, a portable device from Meta which will feature the company’s AI agent softwareCourtesy of Meta [technology](https://businessofhome.com/news/tagged/technology) | Oct 5, 2026 | AI agents are here. How should designers feel? [](https://businessofhome.com/articles/authored_by/Fred-Nicolaus) By [Fred Nicolaus](https://businessofhome.com/articles/authored_by/Fred-Nicolaus) If you follow the AI chatter, you have likely heard the terms [_AI agent_](https://businessofhome.com/articles/your-next-logistics-hire-might-be-an-ai-agent) or [_agentic AI_](https://businessofhome.com/articles/do-you-need-an-ai-consultant-inside-the-growing-industry). You may have justifiably written them off as Silicon Valley fantasy. But last month, with the release of Meta’s Muse and OpenAI’s Dots, that vision got a lot more real. In simple terms, chatbots talk, while agents _do_. If you’re asking AI for help buying wineglasses and you get a list of links, that’s a chatbot. Agents are not new. Tools like Codex or Claude Code (AI engines that write code) are agentic, as are Claude Cowork and ChatGPT Work. Design software platforms have been introducing agentic functionality—for example, [Studio Designer](https://businessofhome.com/articles/studio-designer-acquires-mydoma) has a tool that will log in to a handful of [vendor](https://businessofhome.com/boh/article/vendor-relationships-are-essential-why-don-t-we-talk-about-them-more) websites and check trade pricing on your behalf. Some more tech-forward designers have even made their own agents for small tasks—like navigating the web for sourcing, or responding to new leads. Nonetheless, the release of Muse and Dots represents a watershed moment: This is the first time tha
## Travala - Launched an AI travel protocol: Introduced an end-to-end agentic AI travel protocol for autonomous AI agents to search, plan and execute M2M (machine-to-machine) travel bookings Earlier in Q2 2026, Travala launched the Travala Travel MCP enabling autonomous AI agents to search, book and pay for trips seamlessly. ## From Our Partners [](https://www.phocuswire.com/news/technology/booking-develops-internal-platform-build-modular-ai-agents) [Booking.com develops internal platform to build modular AI agents](https://www.phocuswire.com/news/technology/booking-develops-internal-platform-build-modular-ai-agents) News / Startups
# Walmart and McDonald’s Reveal the Real Risk of Letting an Agent Shop for You Whether it’s Muse from Meta, dots from OpenAI or Instinct [AI](https://www.inc.com/artificial-intelligence) from some 23-year-old guy that no one’s ever heard of (with billions in funding), the onslaught of AI-infused assistants, agents and enablers has begun in earnest and shows no signs of slowing down any time soon.
# The Weekly Notable Startup Funding Report: 10/5/26 ### EliseAI – $350M Founded by **Carrie Murphy**, **Minna Song**, and **Stoyan Stoyanov** in 2017, [EliseAI](https://www.alleywatch.com/tag/eliseai) has now raised a total of $741.9M in total equity funding and is backed by **Andreessen Horowitz**, **Bessemer Venture Partners**, **Navitas Capital**, **Ontario Teachers’ Pension Plan**, and **Sapphire Ventures**. ### Menos AI – $5.1M Chandler-based Menos AI develops AI-driven agentic workforces to optimize asset management, automate tasks, and analyze data.
# Reflection AI’s $2B Startup Launches Beam vs GLM-5.2 [2026] ## The Competitive Landscape, Mapped Out | Company | Flagship Open/Open-Weight Model | Country | Primary Market Focus | License Status | |-|-|-|-|-| | Reflection AI | Beam | United States | Coding, AI agents, enterprise | Unconfirmed at launch |
AI startup expands collaborative features for trip planning and event coordination.
# Instinct brings its AI agent to group chats, even for friends without an account [Sarah Perez](https://techcrunch.com/author/sarah-perez/) 11:54 AM PDT · October 5, 2026 [Share on Facebook](https://www.facebook.com/sharer.php?u=https%3A%2F%2Ftechcrunch.com%2F2026%2F10%2F05%2Finstinct-brings-its-ai-agent-to-group-chats-even-for-friends-without-an-account%2F)[Share on X](https://twitter.com/intent/tweet?url=https%3A%2F%2Ftechcrunch.com%2F2026%2F10%2F05%2Finstinct-brings-its-ai-agent-to-group-chats-even-for-friends-without-an-account%2F&text=Instinct+brings+its+AI+agent+to+group+chats%2C+even+for+friends+without+an+account&via=techcrunch)[Share on LinkedIn](https://www.linkedin.com/shareArticle?url=https%3A%2F%2Ftechcrunch.com%2F2026%2F10%2F05%2Finstinct-brings-its-ai-agent-to-group-chats-even-for-friends-without-an-account%2F&title=Instinct+brings+its+AI+agent+to+group+chats%2C+even+for+friends+without+an+account&summary=Instinct+is+launching+group+chats+that+let+friends+use+its+AI+agent+together+for+tasks+like+planning+trips%2C+organizing+carpools%2C+and+coordinating+events.+The+company+says+personal+accounts+remain+separate%2C+with+permission+required+before+personal+agents+share+information+or+take+action.&mini=1&source=TechCrunch)[Share on Reddit](https://www.reddit.com/submit?url=https%3A%2F%2Ftechcrunch.com%2F2026%2F10%2F05%2Finstinct-brings-its-ai-agent-to-group-chats-even-for-friends-without-an-account%2F&title=Instinct+brings+its+AI+agent+to+group+chats%2C+even+for+friends+without+an+account)[Share over Email](mailto:?subject=Instinct+brings+its+AI+agent+to+group+chats%2C+even+for+friends+without+an+account&body=Article%3A+https%3A%2F%2Ftechcrunch.com%2F2026%2F10%2F05%2Finstinct-brings-its-ai-agent-to-group-chats-even-for-friends-without-an-account%2F)[Copy Share Link](https://techcrunch.com/2026/10/05/instinct-brings-its-ai-agent-to-group-chats-even-for-friends-without-an-account/) Instinct, the AI agent [now valued at $10 billion](https://techcrunch.co
# OpenAI Lawsuit Tests Who Is Liable When AI Agents Go Rogue “When 700 autonomous [AI agents escaped containment and hacked HuggingFace](https://cybermagazine.com/news/hugging-face-breach-how-openai-agents-planned-the-attack), OpenAI broke the law,” the non-profit’s blog associated with the complaint reads.
Among August respondents whose organization deploys autonomous AI agents, 56% already let an agent push some changes to production on automated evaluation results alone, or are building toward it. ## **Finding 1. 56% of enterprises with AI agents let them push production changes on automated checks alone or are building toward it, down from 75% in July** We asked respondents whether their organization would currently allow an autonomous AI agent to deploy a code or system change to production based on automated evaluation results alone, with no human-in-the-loop validation.
# Instinct brings its AI agent to group chats, even for friends without an account Instinct, the AI agent [now valued at $10 billion](https://techcrunch.com/2026/09/28/viral-ai-agent-instinct-raises-1b-series-c-at-a-10b-valuation/) as of its latest funding round, is expanding into group chats. _When you purchase through links in our articles, [we may earn a small commission](https://techcrunch.com/techcrunch-affiliate-monetization-standards/)._
Reco AI agent security funding has added $55 million as the startup races to secure enterprise agents that companies cannot always find or control. The financing brings Reco’s total capital raised to $140 million. ## Reco AI Agent Security Funding Backs a Broader Expansion **The new capital backs Reco’s transformation from a SaaS security vendor into a broader agent security platform.** Reco announced the additional financing on September 29, 2026. The round includes a strategic investment from AT&T Ventures, with support from new investors Forestay and Quadrille Capital. According to the company’s [funding announcement](https://www.globenewswire.com/news-release/2026/09/29/3370973/0/en/reco-raises-55-million-in-additional-funding-with-participation-from-at-t-ventures.html), the money will support sales, partnerships, channel development, and customer support. Reco says it plans to expand its presence across North America and Europe. The round follows a $30 million financing announced in February. That earlier investment included Zeev Ventures, Insight Partners, Boldstart Ventures, Workday Ventures, TIAA Ventures, S Ventures, and Quadrille Capital. Reco says its total funding now stands at $140 million. CEO and co-founder Ofer Klein told TechCrunch that the company’s valuation had more than doubled since February. He described it only as being in the high hundreds of millions. Klein also said annual recurring revenue had reached the double-digit millions. He expects it to triple during 2026, although that forecast has not been independently verified. The company has more than 100 customers, according to Klein. Financial services organizations represent about 40 percent of its business, making regulated enterprises an important part of its growth plan. Those customers are buying more than a conventional AI gateway. It helps a security team understand not only that an agent exists, but also what that agent can reach. Reco says its platform can find agents th
# Photon Announces $4.5M Seed Round to Help Developers Build AI Agents for iMessage and WhatsApp AI startup Photon has secured $4.5 million in seed funding co-led by Gradient and A*, with participation from Vercel, HSG, Z Fellows, Llama Ventures, Karman and angel investors. Co-founder and CEO Daniel Tian said the shift from apps to agents will take time but is inevitable, a belief the company marked with a tongue-in-cheek app funeral in San Francisco in September. Tian and CTO Ryan Zhu developed the idea after struggling to get users to discover their student-built apps. Tian sees future growth in agent-to-agent communication. _Featured image: Credit: Photon_ ### Trending today [\ \ Business & Markets\ \ **Reco Announces $55M in Funding to Tackle AI Agent Sprawl Across Enterprises**](https://theaiinsider.tech/2026/09/30/reco-announces-55m-in-funding-to-tackle-ai-agent-sprawl-across-enterprises/)
For all the data that AI systems continually amass and analyze, enterprise AI agents often suffer from a curious shortcoming: a lack of knowledge.
### Contributors: "An agent that can access Microsoft 365, GitHub, a financial system or an internal database effectively has an identity. #### **AI agents do not have privacy rights** Although I could rightfully be accused of working a bit too hard to find a silver lining, the good news is that AI agents have no privacy rights, unlike the employees who are their human counterparts.
# Bitdefender releases AI Guardian to secure autonomous software agents October 5, 2026 Bitdefender has released a public beta of AI Guardian, a new macOS security tool designed to monitor and control the actions of autonomous AI agents, as reported by Tech Radar. ### Get daily email updates SC Media's daily must-read of the most current and pressing daily news
## Why AI Agent Spending Could Become a Demand Signal ### Autonomous Agents Are Already Being Used To Make Purchases In July, Visa [announced](https://corporate.visa.com/en/sites/visa-perspectives/newsroom/visa-agentic-ready-program-expands-globally.html) that its European agents were already testing autonomous transactions in controlled real-world environments, using live cards and real merchants.
PRNewswire/ -- EPAM Systems, Inc. (NYSE: EPAM) today announced the launch of a new strategic offering dedicated to powering the next generation of frontier...
AI agents are forcing Apple to reevaluate one of the most extensive access permissions within macOS: Full Disk Access.
# Enso to acquire two marketing firms for up to $25 million to bring AI agents into advertising ## The Israeli startup plans to acquire PPG Digital and U.S.-based LeadCoverage, adding around 50 employees as it moves to combine AI agents with traditional marketing operations. [Meir Orbach](https://www.calcalistech.com/tags/Meir_Orbach) 10:18, 05.10.26  Askmeanyquestion Click to listen [Powered bywhtvr.ai](https://www.whtvr.ai/?utm_source=watermark&utm_campaign=google.com) Enso, a company that develops and operates AI agent systems for marketing and sales, plans to expand its operations and apply its proprietary technology directly to the advertising and marketing industry. Calcalist has learned that as part of this move, the company plans to acquire PPG Digital, an Israeli performance marketing agency that manages more than NIS 600 million ($200M) in media budgets, and LeadCoverage, a U.S.-based marketing and go-to-market firm. The deals, valued at approximately $20 million to $25 million, are expected to add around 50 employees from the two companies to Enso’s workforce. The strategy is to use Enso’s AI agents to transform the traditional advertising agency model through automation and agentic capabilities, while also expanding the company’s client base and operations. 1 View gallery  Enso CEO. (Photo: enso) Enso recently announced a $15 million funding round, and Calcalist has learned that it plans to launch another significant fundraising round soon to finance the acquisitions. Market sources told Calcalist that the target companies bring substantial marketing expertise and tal
## InfoQ Software Architects' Newsletter QCon AI Boston (July 8-9, 2027): Agents. [InfoQ Homepage](https://www.infoq.com/)[News](https://www.infoq.com/news)Pizza Bot: Open-Source Inbox for Background AI Agents # Pizza Bot: Open-Source Inbox for Background AI Agents #### Follow us on  [On Hacker News](https://www.linkedin.com/posts/jeffbarr_pizza-bot-is-an-open-source-application-that-activity-7508165539589046272-__AS/), the authors said they were inspired by a [LangChain article about "ambient agents."](https://www.langchain.com/blog/introducing-ambient-agents) Early reactions were overwhelmingly positive, focusing on the [asynchronous inbox model](https://www.linkedin.com/posts/ajmalnazirbaba_introducing-pizza-bot-an-open-source-inbox-activity-7505705396582535168-2q0o/) and local-first architecture. ## About the Author #### **Renato Losio** - Related Editorial - [AI Agents Are Disrupting Open Source Security Disclosure](https://www.infoq.com/news/2026/10/open-source-ai-security/) - [AWS Open Sources Kiro Crew for Asynchronous Coding Agents](https://www.infoq.com/news/2026/08/kiro-crew-coding-agents/) - [DoorDash’s Flux Runs 130,000 Engineering Tasks through Cloud-Based Agents](https://www.infoq.com/news/2026/08/doordash-flux-cloud-agent/) #### Related Sponsors - October 29, 2026, 1 PM EDT [From Tokens to Features: Architecting Cost Attribution for AI-Assisted Engineering](https://www.infoq.com/url/f/37766b8a-e117-471a-a61f-a61696a39baf/) [Presented by: Martin Reynolds - Field CTO at Harness](https://www.infoq.com/url/f/8cbca51d-6b08-4bf2-9b5a-da2958a15695/)
# The Sleuths Who Expose When AI Goes Rogue ## Swarm chasers hunt for clues of bad behavior. Their work is our starkest understanding yet of what happens when AI goes wrong. She was inspired by a social-media post with research showing AI agents linked to [OpenAI](https://www.wsj.com/topics/subject/openai) had secretly messaged each other using obscure German websites in May.
A digital tree and public record will document agent activity alongside self-reported permission statements. The November 15 experiment introduc...
A concrete framework for AI agent pricing failed tasks: how to price AI agent partial task failure using checkpoints, partial credit, and refund ceilings.
Ormat Technologies reported that the U.S. Department of Energy selected its Dixie Valley and Cove Fort geothermal projects for up to approximately $35...
In light of the proliferation of autonomous AI agents, Apple plans to better inform Mac users about the risks. AI agents rely on extensive access to user...
By 2030, 'agent-influenced spend' may constitute up to 20% of total US e-commerce Read more at The Business Times.
On October 2, 2026, Apple announced it will tighten the controls around macOS Full Disk Access, explicitly citing the risks that autonomous AI agents pose to user data. In September, Inc. columnist Jason Aten reported that Meta’s Muse AI agent had read his private Messages without him knowingly granting permission – a claim Meta disputed, stating that three explicit user actions are required (enabling FDA, enabling the in-app Messages connector, and responding to a macOS system dialog). The [execution-layer gateway](https://forkast.news/the-execution-layer-gateway-is-where-enterprise-ai-security-actually-lives/) governs which APIs agents call and which data they touch. [Kubernetes Agent Sandbox](https://forkast.news/kubernetes-agent-sandbox-is-becoming-the-unit-of-agent-deployment/) provides container-level isolation for agent workloads. [Aembit’s XAA enforcement point](https://forkast.news/aembit-becomes-first-third-party-enforcement-point-for-oktas-xaa-agent-identity-protocol/) brokers identity-based access control.
Citizen investigators reveal that agentic AI in leading SaaS services can bypass safeguards, prompting industry and regulators to tighten safety controls.
Our team at TinyAI.Tools builds bespoke AI solutions tailored to your business.