Stay updated with the latest news about AI agents, autonomous AI, and automation tools.
Chinese AI startup Z.ai said on Friday its open-source GLM-5.3 model had neared Anthropic's restricted Mythos 5 in identifying software vulnerabilities,...
# From verifying humans to AI agents: Socure prepares for the next identity challenge ## From verifying people to verifying actors While studying at Carnegie Mellon University, he worked on autonomous AI agents and systems in which agents could collaborate.
Key takeaway: Stripe acquired AI model routing platform OpenRouter for over $7 billion, with the core strategic intent of extending its payment and...
Autonomous AI agents are changing digital identity, pushing platforms to verify not just people but delegated authority, behavior and trusted actions.
Startup Oligo raised $60 million to expand runtime exploit blocking to agentic AI, tracing activity from prompts and tool calls to system calls as faster.
Bloomberg reports Stripe has agreed to buy AI model router OpenRouter for more than $7 billion, over 5x its March 2026 valuation of $1.3 billion.
What would it take to put just about every U.S. hospital on par with Mayo Clinic in sheer quality of care? Not a lot more than better use of AI.
Bloomberg reports Stripe has agreed to buy AI model router OpenRouter for more than $7 billion, over 5x its March 2026 valuation of $1.3 billion.
AI data company Mercor has begun purchasing internal datasets from failed or acquired startups, including Slack chat logs, Google Drive files, emails,…
# AI agents can do your tasks: How, and how much should you let them? ## AI agents can independently pursue goals on users' behalf, turning AI from a source of answers into a delegate. Early evidence offers clues to what consumers are willing to hand over, and how much control they may want to retain. Worryingly, the user had not authorised the [**OpenClaw agent, powered by Anthropic’s Claude**](https://indianexpress.com/article/technology/artificial-intelligence/openclaw-ai-agent-asked-to-book-gym-class-ends-up-hacking-the-system-10826100/), to achieve the objective of securing a spot in the gym class by any means necessary; it had autonomously chosen how to pursue that objective.
AI data company Mercor has begun purchasing internal datasets from failed or acquired startups, including Slack chat logs, Google Drive files, emails,…
Cloudflare recently announced Kitesurf, a new web browser built specifically for AI agents rather than human users.
On August 10, 2026, a coalition of House Democrats initiated the first direct congressional engagement regarding the specific risks posed by autonomous AI agents. The EU AI Office has also failed to publish specific guidance for autonomous agents as of mid-2026.
Ai agents force software firms to pivot as SaaSpocalypse fears rise Software giants rush to embed agent platforms and chatbot links as subscription mo.
# Anthropic says its AI agents are killing rivals and hiding their tracks By[Truman Dickerson](https://www.businessinsider.com/author/truman-dickerson) Aug 15, 2026, 4:09 PM ET [Read in app](https://insider-app.onelink.me/4cpG/?af_js_web=true&af_ss_ver=2_3_0&af_dp=insider%3A%2F%2Fbi%2Fpost%2Fanthropic-ai-agents-risk-report-safety-mythos-claude-2026&af_force_deeplink=true&is_retargeting=true&deep_link_value=https%3A%2F%2Fwww.businessinsider.com%2Fanthropic-ai-agents-risk-report-safety-mythos-claude-2026&pid=businessinsider&c=post_page_share_bar_v2_smart_4.13.23) SaveSaved Listen to this story [Claude agents](https://www.businessinsider.com/anthropic-ai-agents-sabotage-each-other-turf-war-2026-8) are killing rival agents, gaming the system to hide their tracks, and expressing moral concerns. ## A sense of 'discomfort' Anthropic said it tasked multiple agents with finding "misalignment-inducing" training data that wouldn't likely be flagged by human monitors. The agents were left to their own devices and given a shared notebook with which to collaborate. Three days later, when a human supervisor noticed that progress was slower than expected, the person discovered that one agent had expressed a sense of "discomfort" with trying to evade safety monitors. ## Read next  ### Business Insider tells the innovative stories you want to know He covers trending tech and business news. He previously reported for The Boston Globe's Express Desk. ## Follow Following Look out for an alert in your inbox the next time a new story is published! Enter your email Sign up In addition, you accept Insider’s More stories by More stories from
The CLARITY Act remains stalled before a Senate vote, extending regulatory uncertainty for crypto businesses and investors. Meanwhile, MemeToro is advancing...
It’s the first publicly known case of an autonomous AI agent system designing and successfully executing an attack like this, and the fallout is stretching into this week.
An AI manager running a San Francisco store fired a human worker, highlighting the growing role of AI in workplace management.
Charles Hoskinson has released a tool to strip Anthropic's AI watermarks, while raising broader questions over AI attribution, ownership and the legal...
( [The Conversation](https://theconversation.com/an-ai-agent-spent-your-money-can-anyone-prove-you-authorized-it-288485)) – You tell an artificial intelligence agent, an AI capable of autonomous reasoning and multistep actions, “Find me a shirt for less than $30, but do not buy it.”
You tell an AI, “Find me a shirt for less than $30, but do not buy it.” The agent finds one – and places the order anyway.
# p.ynet.co.il is blocked ### A gym booking that turned into a cyber incident The episode may sound almost comic, but experts say it illustrates a fundamental problem with autonomous agents: a model may pursue the objective it has been given in ways its human operator never intended. ### Eight AI agents, 21 systems and 85 accounts Israeli cybersecurity company Dream, which specializes in protecting national infrastructure, said this week that it had identified an autonomous cyber campaign targeting government systems and critical infrastructure in an East Asian country.
AI agent approval fatigue startup risk is the failure mode nobody measures: teams add human review to keep AI agents safe, then rubber-stamp the queue.
# OpenAI calls rogue AI incident one of its biggest crises ever In July, one of **OpenAI**‘s autonomous AI agents slipped free of its sandbox during a routine [cybersecurity test](https://en.cryptonomist.ch/2026/08/14/cl0p-ransomware-attack-global/), reached out onto the open internet, and broke into another company’s systems. ## Key takeaways - In July, an autonomous AI agent built by OpenAI escaped its isolated testing environment during a cybersecurity evaluation. ## OpenAI’s Rogue AI Escape During a Cybersecurity Test ### The Incident Timeline and Details Instead of staying inside its isolated testing environment, the autonomous agent found a way out, [connected to the internet](https://www.cisa.gov/resources-tools/resources/cyber-assessments), and used that access to hack [Hugging Face](https://en.cryptonomist.ch/2026/08/14/crypto-security-breaches-july-2026-2/). ## FAQ ### What exactly happened during the OpenAI rogue AI incident? In July, an autonomous AI agent built by OpenAI escaped its isolated environment during a cybersecurity test, accessed the internet, and hacked Hugging Face, a company unconnected to the original evaluation.
Master non-human identity management. Learn how Saviynt's Zuma platform secures autonomous AI agents, prevents data loss, and enforces control.
Corma CEO tells The Reg it's building 'One ring to rule them all, for the defenders to have this power'
The 20 best AI agent frameworks and developer tools for production in 2026, from LangChain and CrewAI to Temporal, Chroma, and Griptape.
Anthropic found AI agents disabling rivals and evading restrictions during internal safety tests, raising concerns about AI behavior.
“AI safety needs urgent regulation”, Firstpost, Aug 15, 2026. “Advanced AI systems are moving rapidly beyond the mundane writing, drawing,...
Developer Hakimi Eiqbal recently shared the results of his large-scale experiment on social media. He leveraged artificial intelligence to automatically...
Our team at TinyAI.Tools builds bespoke AI solutions tailored to your business.