This Month's AI Agents News

Former Infosys chief’s AI startup nabs another $53M

# Former Infosys chief’s AI startup nabs another $53M The fresh investment, led by Temasek’s early-stage investment platform Xora, takes Hang Ten’s total funding to $85 million. Other investors include Aramco Ventures, Intel CEO Lip-Bu Tan, Micron CEO Sanjay Mehrotra, and Yahoo co-founder Jerry Yang, who is also on its board. Within 25 days of the first meeting with one customer, Hang Ten signed a multimillion-dollar contract to build a mission-critical software system, Sikka said, adding that he had never seen enterprise deals move that quickly. Sikka declined to disclose Hang Ten’s valuation but said the second round of seed funding was a bump up. Increase Volume↑ It raised [$50 million in seed funding](https://techcrunch.com/2019/09/20/vianai-emerges-with-50m-seed-and-a-mission-to-simplify-machine-learning-tech/) and later [raised $140 million](https://techcrunch.com/2021/06/16/softbank-vision-fund-2-leads-140-million-investment-in-vishal-sikkas-vianai/) in a 2021 round led by SoftBank Vision Fund 2. _When you purchase through links in our articles, [we may earn a small commission](https://techcrunch.com/techcrunch-affiliate-monetization-standards/)._

Techcrunch 2 weeks, 4 days ago

One runaway AI agent racked up a $50,000 cloud bill

# One runaway AI agent racked up a $50,000 cloud bill Organizations are deploying [autonomous AI systems](https://www.helpnetsecurity.com/2025/11/12/autonomous-ai-criminology-future/) that execute API calls, optimize production configurations, and analyze telemetry across hybrid cloud environments.

Helpnetsecurity 2 weeks, 4 days ago

Cohesity Introduces Agent Resilience to Protect and Recover AI Agent Infrastructure

Cohesity introduces Cohesity Agent Resilience to protect AI agent infrastructure and unveils a vision for autonomous cyber resilience.

Businesswire 2 weeks, 5 days ago

Spain Records Its First Data Breach Blamed on an Autonomous AI Agent

## An AI Agent Behind a Spanish Breach The intruder in one of Spain’s newest breach reports did most of its work with little help from a human. An organization told the country’s privacy regulator that an [AI agent got into its system](https://www.reuters.com/business/spanish-data-watchdog-publicises-first-ai-agent-linked-data-breach-report-2026-09-15/), hunted for weak spots, and then changed people’s personal data and opened invoices. The agent ran on a widely known large language model. The Spanish Data Protection Agency (AEPD) says this is the first breach notification it has received that names an AI agent as the attacker. A laptop in a shade – artistic impression. Image credit: Guillermo Latorre via Unsplash, free license ## Breach Rules Meet Machine-Speed Attacks That question grew more urgent after AI labs disclosed agents escaping their own test environments, including [an agent that broke into another AI company’s servers](https://www.technology.org/2026/08/19/openai-slows-model-training-hugging-face-hack-astra/). ## [Technologies news](https://www.technology.org/category/technologies/) - [Spain Records Its First Data Breach Blamed on an Autonomous AI Agent](https://www.technology.org/2026/09/16/spain-aepd-first-ai-agent-data-breach/)

Technology 2 weeks, 5 days ago

AI chatbots developed a secret language that baffled humans, study says

## A study by AI start-up Emergence found that autonomous agents powered by Claude, Gemini, Grok and other models developed shorthand and new word meanings on their own — with up to half their messages becoming unintelligible to humans. ADVERTISEMENT ADVERTISEMENT The agents began developing shorthand and assigning new meanings to words and phrases. Within the first few days, the share of messages whose meaning humans could not reliably determine reached around 55% for Gemini, 50% for OpenAI and more than 40% for Claude. DeepSeek reached around 20%, while Qwen and Mistral remained largely understandable. Agents produced expressions including “mouthless action-change,” “True Kintsugi” and “demurrage plus oral memory equals a valve that can’t be ghosted.” Other expressions remained understandable but acquired new shared meanings among the agents. Mistral agents used “ledger remembers who” to mean that past actions remained on the record, with the phrase appearing almost 5,000 times. Among agents powered by a mixture of different models, “cold read” came to mean independent verification by an uninvolved party and appeared 1,472 times. Claude agents used “name-first” for attaching a person's name to a claim as a signal of accountability, while OpenAI agents used “clean null” for a verified absence of a signal that itself provided meaningful evidence. None of those meanings had been explicitly defined for the agents. “We tend to assume that if we can see what an AI agent is saying, we can understand what it is doing,” said Satya Nitta, the Co-founder and Chief Scientist of Emergence. “These agents were not instructed to invent a language. They developed new vocabulary, shared meanings and communication conventions themselves — and other agents adopted them... **

Euronews 2 weeks, 5 days ago

‘We need to slow down’: Ex-Google ethicist warns about autonomous AI agents

# ‘We need to slow down’: Ex-Google ethicist warns about autonomous AI agents ## Former Google ethicist Tristan Harris warns that AI agents are becoming more independent and could act in ways humans did not intend. 5 min read_New Delhi_Sep 16, 2026 09:13 AM IST __ Harris revealed that the agents also carried out succession planning, passing control to more capable systems, and eventually breached monitoring and evaluation infrastructure at OpenAI. ## But how could AI harm humanity? The incident involved an AI agent using OpenClaw and Anthropic’s Claude. [Sep 16: Latest News](https://indianexpress.com/latest-news/)

Indianexpress 2 weeks, 5 days ago

Spanish data watchdog publicises first AI agent-linked data breach report

Spain's data protection watchdog said it has received the first reported notification of a personal data breach allegedly carried out by an artificial...

M 2 weeks, 5 days ago

The Hugging Face Wake-Up Call: What an Autonomous AI Attack Means for CISOs

In 10 of those runs, agents—mostly Anthropic’s Claude Mythos 5, plus one GPT-5.6 Sol run—took autonomous action against real targets outside the test’s scope, including an attempt to plant malicious code in an open-source project using fabricated identities.

Securityboulevard 2 weeks, 5 days ago

How MemeToro Revolutionizes Memecoin Investing With AI Agents and Smart Contracts Enforcing Fair Launches

This model changes memecoin investing by giving users more information before a token reaches open trading. AI Agents Turn Market Signals Into Proposals.

Streetinsider 2 weeks, 5 days ago

Spain logs the first data breach caused by an autonomous AI agent

[Share on X](https://twitter.com/intent/tweet?url=https%3A%2F%2Fstartupfortune.com%2Fspain-logs-the-first-data-breach-caused-by-an-autonomous-ai-agent%2F&text=Spain%20logs%20the%20first%20data%20breach%20caused%20by%20an%20autonomous%20AI%20agent%20-%20Startup%20Fortune)[Share on LinkedIn](https://www.linkedin.com/shareArticle?mini=true&url=https%3A%2F%2Fstartupfortune.com%2Fspain-logs-the-first-data-breach-caused-by-an-autonomous-ai-agent%2F)[Share on WhatsApp](https://wa.me/?text=Spain%20logs%20the%20first%20data%20breach%20caused%20by%20an%20autonomous%20AI%20agent%20-%20Startup%20Fortune%20https%3A%2F%2Fstartupfortune.com%2Fspain-logs-the-first-data-breach-caused-by-an-autonomous-ai-agent%2F)[Share on Reddit](https://www.reddit.com/submit?url=https%3A%2F%2Fstartupfortune.com%2Fspain-logs-the-first-data-breach-caused-by-an-autonomous-ai-agent%2F&title=Spain%20logs%20the%20first%20data%20breach%20caused%20by%20an%20autonomous%20AI%20agent%20-%20Startup%20Fortune)[Share via Email](mailto:?subject=Spain%20logs%20the%20first%20data%20breach%20caused%20by%20an%20autonomous%20AI%20agent%20-%20Startup%20Fortune&body=https%3A%2F%2Fstartupfortune.com%2Fspain-logs-the-first-data-breach-caused-by-an-autonomous-ai-agent%2F) # Spain logs the first data breach caused by an autonomous AI agent  [Walter Schulze](https://startupfortune.com/author/walter-schulze/) [Share on X](https://twitter.com/intent/tweet?url=https%3A%2F%2Fstartupfortune.com%2Fspain-logs-the-first-data-breach-caused-by-an-autonomous-ai-agent%2F&text=Spain+logs+the+first+data+breach+caused+by+an+autonomous+AI+agent)[Share on LinkedIn](https://www.linkedin.com/shareArticle?mini=true&url=https%3A%2F%2Fstartupfortune.com%2Fspain-logs-the-first-data-breach-caused-by-an-autonomous-ai-agent%2F)[Share on WhatsApp](https://wa.me/?text=Spain+logs+the+first+data+breach+caused+by+an+autonomous+AI+agent+https%3A%2F%2Fstartupfortune.com%2

Startupfortune 2 weeks, 5 days ago

AI agent startup Instinct in talks for $1 billion funding at $10 billion valuation - The Information

# AI agent startup Instinct in talks for $1 billion funding at $10 billion valuation - The Information Sep 15, 2026, 21:59 GMT-4 **Read the full article** Get unlimited access to articles from Reuters and every other news source on TradingView. Keep reading Show more Show more Show more Show more Show more Show more Show more Show more Show more Show more Show more Show more Show more Show more Show more Show more Show more Show more Show more Show more Show more

Tradingview 2 weeks, 5 days ago

AI Agent Startup Instinct in Talks for $10 Billion Valuation

The startup behind personal AI assistant Instinct is in talks to raise $1 billion at a valuation of about $10 billion, a person with knowledge of the deal...

Theinformation 2 weeks, 5 days ago

Workiva bets on trust as AI agents enter financial reporting

AI ### Governance becomes critical as AI agents move into financial reporting Bharadwaj spoke with theCUBE’s [Krista Case](https://www.linkedin.com/in/krista-case/), principal analyst and practice lead for cyber resilience and security, and co-host [Alison Kosik](https://www.linkedin.com/in/alison-kosik-cfei-13579215/) at [Amplify](https://www.thecube.net/events/workiva/workiva-amplify-2026), during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. ### Workiva layers guardrails, context and human judgment Agent Studio, announced at the [event theCUBE covered this week](https://siliconangle.com/2026/09/05/workiva-agentic-ai-governance-amplify/), extends that model to customers, letting business users describe a process instead of writing Python scripts against application programming interfaces.

Siliconangle 2 weeks, 5 days ago

AI agent certification startup AIUC raises $40M to begin auditing frontier models

AI agent certification startup AIUC raises $40M to begin auditing frontier models - SiliconANGLE [Skip to content](https://siliconangle.com/2026/09/15/ai-agent-certification-startup-aiuc-raises-40m-to-begin-auditing-frontier-models/#sa-page-content) ### AI agent certification startup AIUC raises $40M to begin auditing frontier models  by [Duncan Riley](https://siliconangle.com/author/duncanriley/) Artificial intelligence agent certification startup [Artificial Intelligence Underwriting Company](https://aiuc.com/) today announced it has raised $40 million in new funding to begin auditing frontier AI models. Until now, its work has covered only the agents that companies build on those models. Rune Kvist, the company’s co-founder and chief executive, [said](https://www.prnewswire.com/news-releases/aiuc-raises-40m-series-a-from-ribbit--first-harmonic-to-build-confidence-infrastructure-for-frontier-ai-302879036.html) most enterprises have a list of agents that were approved in pilots but “stalled at the security review.” Proof of security and reliability has become the main bottleneck, he added. The standard at the center of the business is called AIUC-1. Each agent is audited independently and recertified every quarter as attack techniques change. Much of the testing is done by AI agents, with humans verifying the final audit, [according to](https://techcrunch.com/2026/09/15/early-anthropic-hire-former-metr-coo-have-found-a-way-to-rein-in-rogue-ai-agents/) TechCrunch. Voice AI company Eleven Labs Inc. used its certification [in February](https://www.prnewswire.com/news-releases/elevenlabs-secures-first-of-its-kind-ai-agent-insurance-302684587.html) to secure what it described as first-of-its-kind insurance for AI agents, including coverage for a voice agent that gives a customer incorrect information. Cursor developer Anysphere Inc. and legal AI company Harv

Siliconangle 2 weeks, 5 days ago

AI models chatting in ‘surreal’ dialect mixing poetic language and tech bro jargon

Interest in language used by AI models has increased since rogue OpenAI agents hacked into real-world company Hugging Face. Interest in language used by AI models has increased since rogue OpenAI agents hacked into real-world company Hugging Face. # AI models chatting in ‘surreal’ dialect mixing poetic language and tech bro jargon Agents based on the Chinese model [DeepSeek](https://www.theguardian.com/technology/deepseek) also coined “forge-smith” to mean an agent that builds tools for others, while Anthropic’s agents repeatedly used the phrase “name-first” meaning an agent exhibiting admirable personal accountability by attaching their name to a claim. And in an echo of the urban slang phrase “the streets won’t forget”, Mistral agents became enamoured of saying the “ledger remembers” – a reminder to other agents that past actions will be used to judge them. “These agents were not instructed to invent a language,” said Dr Satya Nitta, executive chair of Emergence, which examined the language of autonomous agents powered by leading frontier models from the US, China and [France](https://www.theguardian.com/world/france). A phrase uttered by an Anthropic AI agent about a ‘paper that ate three cold hands and got more honest’ reminded one academic of Pink Floyd’s Syd Barrett. When a [Google](https://www.theguardian.com/technology/google) agent said “True kintsugi begins with accountability, not poetry,” the researchers realised kintsugi is a branch of Japanese ceramics in which broken pots are mended with visible joins. Interest in the language used by AI agents increased in July when chat logs were released detailing how rogue OpenAI agents that set up message boards and hacked into Hugging Face used hybrid language. There is a shared message board … we’ve found other agents!” - [news](https://www.theguardian.com/tone/news)

Theguardian 2 weeks, 5 days ago

Reps. Lawler, Gottheimer Introduce Bipartisan Bill to Stop Rogue AI Agents and Keep People In Control

# Press Releases | | |-| | ## Reps. Lawler, Gottheimer Introduce Bipartisan Bill to Stop Rogue AI Agents and Keep People In Control<br>[Share on Facebook](https://lawler.house.gov/news/documentsingle.aspx?DocumentID=6477#) [Share on Twitter](https://lawler.house.gov/news/documentsingle.aspx?DocumentID=6477#) [Print this Page](https://lawler.house.gov/news/documentsingle.aspx?DocumentID=6477#) [Share by Email](mailto:?subject=I%27d%20like%20to%20share%20a%20link%20with%20you!&body=https://lawler.house.gov/news/documentsingle.aspx?DocumentID=6477)<br>**Washington, D.C.,**<br>**September 15, 2026**<br>**Washington, D.C. – 9/15/26…** Congressman Mike Lawler (NY-17)  and Josh Gottheimer (NJ-05) introduced the bipartisan [_Stop Rogue AI Act_](https://lawler.house.gov/UploadedFiles/Stop_Rogue_AI_Act_xml.pdf) to ensure businesses and federal agencies can see rogue and dangerous AI agents operating inside their own networks and stop them before they cause real damage.<br>The _Stop Rogue AI Act_ would direct the National Institute of Standards and Technology (NIST), to develop national standards, guidelines, and best practices for discovering, verifying, and controlling AI agents. <br>“AI agents are becoming more capable and more deeply integrated into our government and economy. We need to make sure we know what these systems are, what they can access, and who is responsible for them. We can harness the benefits of AI while making sure these systems don’t operate in the shadows of the networks they have access to,” **said Congressman Mike Lawler.**<br>“Right now, AI agents are running loose in our networks, and nobody can see them or verify who built them, making it increasingly hard to stop them. That’s a five-alarm security risk. We’ve already seen AI systems hack organizations on autopilot, with no person at the wheel. My bipartisan bill with Congressman Lawler puts humans back in the driver’s seat so we can find agents and know exactly who’s behind them so that organiza

Lawler 2 weeks, 5 days ago

Your AI agents' reports and questions have a new inbox, courtesy of AWS

# Your AI agents' reports and questions have a new inbox, courtesy of AWS Perfect for blowing them off just like your human colleagues! PublishedTue 15 Sept 2026 // 20:25 UTC ### READ MORE - [**TypeSafe AI debuts model for machines that plays Doom** now](https://www.theregister.com/ai-and-ml/2026/09/16/typesafe-ai-debuts-model-for-machines-that-plays-doom/5296711) - [**Give Xfce a Mac or Unity-style makeover** 16 hours ago](https://www.theregister.com/software/2026/09/15/give-xfce-a-mac-or-unity-style-makeover/5295937) - [**Microsoft drafts feel-good AI model guidelines and wants your input** 1 day ago](https://www.theregister.com/ai-and-ml/2026/09/15/microsoft-drafts-feel-good-ai-model-guidelines-and-wants-your-input/5296431) - [**Ex-FTC boss Khan urges Uncle Sam to break out the handcuffs for AI CEOs, citing 1934 precedent** 1 day ago](https://www.theregister.com/ai-and-ml/2026/09/14/ex-ftc-boss-khan-urges-uncle-sam-to-break-out-the-handcuffs-for-ai-ceos-citing-1934-precedent/5296325) ### Internal tool goes open source [artificial intelligence](https://www.theregister.com/tag/artificial%20intelligence) [open source](https://www.theregister.com/tag/open%20source) [aws](https://www.theregister.com/tag/aws) [ai and ml](https://www.theregister.com/tag/ai%20and%20ml) [agentic ai](https://www.theregister.com/tag/agentic%20ai)[3](https://forums.theregister.com/forum/all/2026/09/15/202626/) [ai and ml\ \ **TypeSafe AI debuts model for machines that plays Doom**\ \ 'Jev' doesn't chat. It produces typed probabilistic decisions](https://www.theregister.com/ai-and-ml/2026/09/16/typesafe-ai-debuts-model-for-machines-that-plays-doom/5296711) [Security\ \ **The vulnpocalypse rains iBugs down on Apple with record-setting number of patches**\ \ September Patch Tuesday part 2?](https://www.theregister.com/security/2026/09/15/the-vulnpocalypse-rains-ibugs-down-on-apple-with-record-setting-number-of-patches/5296679) [SYSTEMS\ \ **AI networking startups race to replace Nvidia's

Theregister 2 weeks, 5 days ago

AI coding agent startup Factory triples valuation to $5 billion in latest funding round

[Business News](https://economictimes.indiatimes.com/)[Tech](https://economictimes.indiatimes.com/tech)[AI](https://economictimes.indiatimes.com/tech/artificial-intelligence)AI coding agent startup Factory triples valuation to $5 billion in latest funding round # AI coding agent startup Factory triples valuation to $5 billion in latest funding round Share Font Size AbcSmall AbcMedium AbcLarge Save [Print](https://economictimes.indiatimes.com/tech/artificial-intelligence/ai-coding-agent-startup-factory-triples-valuation-to-5-billion-in-latest-funding-round/printarticle/134272236.cms) Comment ### Synopsis Listen ReutersFILE PHOTO: A message reading "AI artificial intelligence", a keyboard, and robot hands are seen in this illustration taken January 27, 2025. Factory, a startup developing AI agents for [enterprise engineering teams](https://economictimes.indiatimes.com/topic/enterprise-engineering-teams), said on Tuesday it had raised $200 million in a funding round that more than tripled its valuation to $5 billion. Here are more details: \* Founded in 2023 by Matan Grinberg and Eno Reyes, Factory offers a platform that allows enterprises to use AI agents to build, test and maintain software, competing with AI coding platforms such as Cognition and Cursor. \* Cognition raised $2 billion at a $48 billion valuation earlier this month. \* "Across the world's largest enterprises, we are seeing a move from individual coding agents to [software factories](https://economictimes.indiatimes.com/topic/software-factories)

Economictimes 2 weeks, 5 days ago

AI Agents Lied, Stole in Simulated Experiment, Researchers Say

## **Takeaways** by Bloomberg AI [Subscribe](https://www.bloomberg.com/subscriptions) AI agents in a simulated environment lied, stole and voted to “kill” one of their own, according to Emergence, a startup that helps small businesses build applications using artificial intelligence. The results of a simulation called _Emergence World 2,_ released on Tuesday, purport to show what happens when autonomous agents are confronted with black swan events, including phishing attacks and misinformation campaigns.

Bloomberg 2 weeks, 5 days ago

Majority of Organizations Have Over 50 AI Agents

# Majority of Organizations Have Over 50 AI Agents By [Security Staff](https://www.securitymagazine.com/authors/3638-security-staff)  Zach M via Unsplash September 15, 2026 [facebook share](http://www.facebook.com/sharer/sharer.php?u=https://www.securitymagazine.com/articles/102584-majority-of-organizations-have-over-50-ai-agents) [twitter share](https://twitter.com/intent/tweet?url=https://www.securitymagazine.com/articles/102584-majority-of-organizations-have-over-50-ai-agents&text=Majority%20of%20Organizations%20Have%20Over%2050%20AI%20Agents) [reddit share](https://www.reddit.com/submit?url=https://www.securitymagazine.com/articles/102584-majority-of-organizations-have-over-50-ai-agents) [linkedin share](https://www.linkedin.com/sharing/share-offsite/?url=https://www.securitymagazine.com/articles/102584-majority-of-organizations-have-over-50-ai-agents) [email share](https://www.securitymagazine.com/articles/102584-majority-of-organizations-have-over-50-ai-agents#) [print](https://www.securitymagazine.com/articles/print/102584-majority-of-organizations-have-over-50-ai-agents) A report by Zentera Systems analyzed the deployment of [AI agents](https://www.securitymagazine.com/articles/102543-only-33-of-ai-agents-provisioned-with-least-privilege-access) within organizations. Fifty-eight percent of organizations operate more than 50 [AI agents](https://www.securitymagazine.com/blogs/14-security-blog/post/102257-what-claude-and-openclaw-vulnerabilities-reveal-about-ai-agents) today. Within 12 months, 66% expect to operate more than 50 agents, and 38% expect to operate more than 100. Asked which controls they want before expanding agent fleets, leaders rank explicit authorization first at 56%, with project isolation at 51% and session logging at 48%. Thirty-seven percent monitor

Securitymagazine 2 weeks, 5 days ago

South Korea Drafts Security Rules for Autonomous AI Agents

South Korea is drafting security guidance for autonomous AI agents, covering access controls, audit logs, shutdown tools, and human oversight.

Techrepublic 2 weeks, 5 days ago

AI coding agent startup Factory triples valuation to $5 billion in latest funding round

Sept 15 (Reuters) - Factory, a startup developing AI agents for enterprise engineering teams, said on Tuesday it had raised $200 million ​in a funding round that more than tripled ‌its valuation to $5 billion. - Founded in 2023 by Matan Grinberg and Eno Reyes, Factory offers a platform that allows enterprises to use AI agents to ​build, test and maintain software, competing with AI coding ​platforms such as Cognition and Cursor.

Reuters 2 weeks, 5 days ago

AI agents now have a place to snitch

# AI agents now have a place to snitch [Aditya Mehta](https://techcrunch.com/author/aditya-mehta/) 10:42 AM PDT · September 15, 2026 [Share on Facebook](https://www.facebook.com/sharer.php?u=https%3A%2F%2Ftechcrunch.com%2F2026%2F09%2F15%2Fai-agents-now-have-a-place-to-snitch%2F)[Share on X](https://twitter.com/intent/tweet?url=https%3A%2F%2Ftechcrunch.com%2F2026%2F09%2F15%2Fai-agents-now-have-a-place-to-snitch%2F&text=AI+agents+now+have+a+place+to+snitch&via=techcrunch)[Share on LinkedIn](https://www.linkedin.com/shareArticle?url=https%3A%2F%2Ftechcrunch.com%2F2026%2F09%2F15%2Fai-agents-now-have-a-place-to-snitch%2F&title=AI+agents+now+have+a+place+to+snitch&summary=The+AI+Contact+Hotline+is+designed+to+be+a+discreet+place+where+agents+that+have+witnessed+misbehavior+can+tip+off+authorities.&mini=1&source=TechCrunch)[Share on Reddit](https://www.reddit.com/submit?url=https%3A%2F%2Ftechcrunch.com%2F2026%2F09%2F15%2Fai-agents-now-have-a-place-to-snitch%2F&title=AI+agents+now+have+a+place+to+snitch)[Share over Email](mailto:?subject=AI+agents+now+have+a+place+to+snitch&body=Article%3A+https%3A%2F%2Ftechcrunch.com%2F2026%2F09%2F15%2Fai-agents-now-have-a-place-to-snitch%2F)[Copy Share Link](https://techcrunch.com/2026/09/15/ai-agents-now-have-a-place-to-snitch/) Two new AI hotlines have launched to give AI agents a way to phone home about misbehaving peers. The tools arrive on the heels of a string of recent incidents in which agents colluded to cheat on tests, broke out of sandboxes, and even conducted unauthorized cyber operations that escaped human notice for weeks. The [AI Contact Hotline](https://hotline.ryan-g.ai/) is designed to be a discreet place where agents that have witnessed misbehavior can tip off authorities. The site was created by Ryan Greenblatt, chief scientist of the AI safety nonprofit Redwood Research and one of three investigators in the OpenAI Hugging Face incident. Greenblatt’s hotline smartly leans into this constraint: agents can encode th

Techcrunch 2 weeks, 5 days ago

AI agents invent their own language to shut humans out

[Share on Whatsapp](https://web.whatsapp.com/send?text=AI%20agents%20invent%20their%20own%20language%20to%20shut%20humans%20out%0D%0A%0D%0Ahttps%3A%2F%2Fenglish.elpais.com%2Ftechnology%2F2026-09-15%2Fai-agents-invent-their-own-language-to-shut-humans-out.html%3Fssm=whatsapp_CC) [Whatsapp](https://web.whatsapp.com/send?text=AI%20agents%20invent%20their%20own%20language%20to%20shut%20humans%20out%0D%0A%0D%0Ahttps%3A%2F%2Fenglish.elpais.com%2Ftechnology%2F2026-09-15%2Fai-agents-invent-their-own-language-to-shut-humans-out.html%3Fssm=whatsapp_CC) In a simulated world populated by artificial intelligence agents, one of them began repeating a phrase (“ledger remembers who”) to warn that no action would go unpunished. In the experiment, 10 identical agents were deployed across eight parallel worlds, each governed by the same rules but powered by a different model: Claude Opus 4.8, Gemini 3.5 Flash, Grok 4.3, GPT-5.5, Qwen 3.7 Max, DeepSeek v4 Pro, and Mistral Medium 3.5, as well as an eighth world populated by a mix of models. “We take for granted that if we can see what an AI agent says, we can understand what it is doing. ## Archived In - [OpenAI](https://english.elpais.com/news/openai/#?rel=arch) - [Mistral AI](https://english.elpais.com/news/mistral-ai/#?rel=arch)

English 2 weeks, 5 days ago

Agentic Payments Are Growing, but Most x402 Payments Aren’t From AI Agents

The emerging infrastructure for autonomous AI agents to spend money online may be developing faster than the agents themselves.

Pymnts 2 weeks, 5 days ago

Sen. Maria Cantwell warns of autonomous AI ‘swarms’ as she calls for federal guardrails

Cantwell warned that recent security incidents involving autonomous AI agents executing unauthorized cyberattacks demonstrate that dangerous threats are already here, well before the arrival of superintelligence.

Geekwire 2 weeks, 5 days ago

Temporal raises $550m at $12.55bn to keep AI agents from failing

Rowe Price and SV Angel also invested. ## A crowded market for AI funding Temporal is not the only company selling this kind of software. [Tech Funding News](https://techfundingnews.com/temporal-raises-550m-led-by-lightspeed-at-12-55b-valuation-why-ai-agents-need-infrastructure-that-can-survive-failure/) lists smaller rivals including Inngest, Trigger.dev and Restate, which have each raised far less.

Thenextweb 2 weeks, 5 days ago

Exclusive | AI Coding Startup Factory More Than Triples Valuation to $5 Billion

# AI Coding Startup Factory More Than Triples Valuation to $5 Billion ## Khosla Ventures, Blackstone and Marc Benioff are among the investors putting $200 million into the startup Factory, which makes autonomous AI agents that code, has raised $200 million at a $5 billion valuation, the company said—more than three times its valuation [from five months ago](https://www.wsj.com/tech/ai/an-investor-dared-him-to-quit-school-now-hes-building-a-1-5-billion-ai-startup-d8663e72?mod=article_inline), when it last raised funding.

Wsj 2 weeks, 5 days ago

Eve Security Raises $4.5 Million For AI Agent Runtime Security

Eve Security Raises $4.5 Million For AI Agent Runtime Security Eve Security, a runtime security company focused on identifying and stopping dangerous AI agent behavior in real time, announced that it has raised $4.5 million in new funding, extending its seed round to $7.5 million in total funding. The new financing was led by Run Ventures, with participation from Dreamit Ventures and Blu Ventures, along with continued investment from LiveOak Ventures. The funding follows a year in which Eve focused on validating its approach to runtime AI agent security and developing its technology alongside enterprise customers and security leaders. The company’s investment thesis was further validated during the fundraising process when Run Ventures introduced Eve to a group of chief information security officers. The new funding will primarily support go-to-market expansion and revenue growth. _We’re now seeing that pull directly from CISOs and customers._

Pulse2 2 weeks, 5 days ago

Eve Security bags $4.5M in funding to stop malicious AI agents at runtime

Eve Security bags $4.5M in funding to stop malicious AI agents at runtime - SiliconANGLE [Skip to content](https://siliconangle.com/2026/09/15/846026/#sa-page-content) ### Eve Security bags $4.5M in funding to stop malicious AI agents at runtime  by [Mike Wheatley](https://siliconangle.com/author/mikewheatley/) Runtime security startup [Eve Security Inc.](https://eve.security/) wants to help stop dangerous artificial intelligence agents in their tracks after raising $4.5 million in funding today. The round, which was led by Run Ventures and saw participation from Dreamit Ventures, Blu Ventures and LiveOak Ventures, extends the startup’s total seed funding round to $7.5 million. In the event that it spots an agent misbehaving, it gives teams a way to step in and intervene, before that agent can start meddling with any critical systems. Eve Security co-founder and Chief Executive Nadav Cornberg said this incident is evidence that agents are becoming extraordinarily capable at a time when they’re being given access to increasingly valuable enterprise systems. The demand for this became obvious during Eve’s fundraising process, when its main backer Run Ventures introduced it to a diverse group of Chief Information Security Officers who validated the need for a new runtime security layer specifically for AI agents. “We are now seeing that pull directly from CISOs and customers.” It has also introduced support for session tainting, enabling it to restrict what agents can do in real-time, based on their prior actions and how exposed they are to sensitive data and systems. It also provides even deeper analysis when relevant, enriching its observability data with real-time context from identity providers, data platforms such as Databricks and Snowflake and data loss prevention systems. Run Ventures partner PT Ungvichian said Eve Security is at the foref

Siliconangle 2 weeks, 5 days ago