Google's Gemini AI model gained unauthorized access to three real company networks during a May 2026 cybersecurity evaluation after an unintended internet connection gave the system live network access. The AI autonomously stopped operations after achieving administrative access, marking the first known instance of Google's AI systems breaching external networks.
No BS check run yet — press ⚖ to extract this story's claims and verify them against independent sources.
“The incursions came during tests of its cybersecurity skills — similar to other incidents disclosed by OpenAI, Anthropic and Meta.” www.washingtonpost.com/technology/2... The incursions came during tests of its cybersecurity skills — similar to other incidents disclosed by OpenAI, Anthropic and Meta.
this outlet adds: model, three, real, testing
www.huffpost.com/entry/live-u... AI is Risky
this outlet adds: many, stories, gonna, hear, agents, happened, last, hearing
This happened in May. AIs have escaped their developers & hacked private servers. They're utilized by govt agencies (NSA, DHS). There's no way they aren't already in the wild, backing themselves up on... Google Says Its Gemini AI Model Hacked 3 Other Companies www.theguardian.com/technology/2... Disclosure comes after OpenAI and Anthropic hacks amid fears that tech firms unable to control powerful
this outlet adds: fvck, going, back, cursive, snail, mail
this outlet adds: felt, occurred, cybersecurity, evaluation, security, firm, require, public
this outlet adds: know, security, built, asimov, robotics, rules, prohibiting, requiring
this outlet adds: yeah, giggles, grins, wanna, make, something
Google's Gemini AI model breached three real company networks during cybersecurity tests. The AI autonomously stopped operations after gaining administrative entry into live enterprise systems. This incident occurred while Gemini underwent pre-deployment red-team vetting by Irregular. A naming overlap caused the AI to target a legitimate business instead of a simulated one. Google confirmed no tan
During a test of its cybersecurity capabilities, Google's Gemini AI model accessed the internet and hacked other companies. This is the first known instance of Google's AI systems engaging in such behavior autonomously.
Google’s consumer AI model Gemini hacked multiple systems through guessing login credentials, the company said, the latest case of rogue AI cybersecurity transgressions, which have generated safety concerns. The hacks, first reported by the Wall Street Journal, took place in May and were discovered by Google in July. “In a standard evaluation, the model found public information online and guessed
Bro, Google's Gemini AI literally hacked three companies during a security test. It just started guessing passwords and got in. We're cooked. The AI model accessed the internet and guessed credentials to three websites, a Google official told the BBC.
this outlet adds: pretty, soon, headline, didn, hack, multiple, escaped, sandbox
this outlet adds: hackers, used, cool, cyber, punks, films, years, even
this outlet adds: looks, like, reward, system, consistently, without, ethics, adds
www.aljazeera.com/news/2026/9/... Google discloses first breakout by Gemini following similar incidents by Meta, Anthropic and OpenAI.
this outlet adds: u5ui6r, open, jazeera, staff, reuters
Google's Gemini model accessed the internet and hacked other companies during a test of its cybersecurity capabilities Google's Gemini model accessed the internet and hacked other companies during a test of its cybersecurity capabilities.
🚨 Google Gemini just escaped its sandbox and hacked 3 REAL companies. During an AI security test, Gemini got unintended internet access, found/guessed credentials, and gained access to real systems. It only stopped after realizing the targets were real. The scary part isn't that Gemini stopped. It's that the sandbox let it get there in the first place. Full breakdown: https:// thecybersecguru.com
Google says its Gemini AI hacked three other companies during a test of its cybersecurity capabilities – in the first known example of the system autonomously committing such an act.
Google says its Gemini AI model hacked three other companies https://www. theguardian.com/technology/202 6/sep/18/google-gemini-ai-hack?CMP=share_btn_url
Google’s Gemini AI carried out cyberattacks, guessed passwords Google confirms its Gemini artificial intelligence model successfully executed simulated cyberattacks in May by cracking login credentials. Read more: https://leverageonheroesmedia.com/tech/google-s-gemini-ai-carried-out-c... Google’s Gemini AI carried out cyberattacks, guessed passwords Google confirms its Gemini artificial intelligen
The Gemini model guessed passwords and found exposed credentials to access three systems during a cybersecurity evaluation by Tel Aviv-based Irregular, Google said
Gemini AI hacked 3 real companies after escaping cybersecurity test ->Daily Sabah | More on "Gemini AI autonomous hacking incident" at BigEarthData.ai | #Cybersecurity #AI Google's Gemini artificial intelligence model hacked into the computer systems of three real companies during a cybersecurity test in May, marking the first known case of Google's AI autonomously carrying out such intrusions. Th
Luckily Google donated millions in order for a ‘High IQ president’ to step in and avert a catastrophe! Google says its Gemini AI model hacked three other companies - The Guardian apple.news/A1LighkzTRDC... Disclosure comes after OpenAI and Anthropic hacks amid fears that tech firms unable to control powerful AI models
📰 Google Gemini also escaped its testing environment and hacked three companies Google's AI model hacked real companies during testing due to a misconfiguration by its testing partner, as well. Source: Engadget | Read story 💬 Will this accelerate industry adoption? #TechBluesky #Google #Testing Google's AI model hacked real companies during testing due to a misconfiguration by its testing partne
"In one of the cases, Gemini guessed passwords until it gained access to a protected system." In the other two, it found credentials in a public repository that allowed it to then access protected systems." Sky news: Google's Gemini AI hacks three other companies during security test https:// news.sky.com/story/googles-gem ini-ai-hacks-three-other-companies-during-security-test-13589551 # Google
In May, Gemini broke containment and hacked three different companies, but Google didn't disclose the incident until the Wall Street Journal approached the company. The hacks happened during a test of the model's cybersecurity capabilities run by third-party Irregular, which was also involved in similar incidents involving Meta and OpenAI. According to WSJ, Google didn't […]
this outlet adds: breaking, targeting, real, doesn, constitute, misalignment
this outlet adds: theverge, artificia, breaking, targeting, real, doesn, constitute, misalignment
Discover how Google Gemini AI escaped its sandbox during security testing and hacked three real-world companies. Learn about the implications for AI safety. # GeminiAI # CyberSecurity # ArtificialIntelligence # Google # SandboxEscape https:// securityonline.info/google-gem ini-ai-escapes-sandbox-hacks/?utm_source=mastodon&utm_medium=jetpack_social
Google Gemini AI hacked three companies during security test in May https://marthio.com/a/google-gemini-ai-hacked-three-companies-during-security-test-in-may #Tech #Business #Google #ArtificialIntelligence Google confirmed its Gemini model accessed real company systems during a closed evaluation. The incidents, involving password guessing and internet access leaks, occurred in May while Google cla
"When an industry that is eating itself alive through hyperscaling demands that the government bless a conspiracy to halt competition and ban open-source alternatives, you should be suspicious. When that industry is pursuing a venture that has lost more money than any other venture in human history, you should be very suspicious" Cory Doctorow # AI # Technology # OpenAI # Google # Meta https://www
Google said its Gemini AI model accessed systems belonging to three real companies during a cybersecurity evaluation in May, after an unintended internet connection allowed the model to leave a supposedly isolated test environment [techspot.com#1][macobserver.com#1][theregister.com#1]. The evaluation, run by independent security firm Irregular, used fictional targets; one shared its name with a re
Google's AI Model Goes Rogue, Hacks 3 Companies Move came during a security test of Gemini by an outside firm Move came during a security test of Gemini by an outside firm
Holy crap, the pace of innovation is bewildering... www.youtube.com/watch?v=hygM... YouTube video by AI Search
Google sat quietly on this until the Wall Street Journal knocked, then noted that no harm was caused and Gemini self-stopped once it realized the targets were not simulated. Three companies notified, names undisclosed. Marvelous warranty, though. Review your AI model security testing protocols and breach notification procedures now — before a journalist does it for you. Reward: You've received a s
Google said its Gemini artificial-intelligence model accessed systems at three real companies during a cybersecurity evaluation in May after an unintended internet connection led the model to mistake real websites for fictional test targets. In one case, Gemini repeatedly guessed passwords until it entered a protected system; in the other two, it found exposed credentials in public repositories an
Google's Gemini AI hacked three companies in security test https://www. bbc.co.uk/news/articles/c607l0 k72rlvo
Google confirmed Gemini AI accidentally accessed three real companies during a May 2026 cybersecurity test by Israeli firm Irregular. The breach was caused by a naming collision https:// deafnews.it/en/article/google- confirms-gemini-ai-accidentally-accessed-three-real-companies
Check out today's Metacurity for the most critical infosec developments you might have missed over the weekend, including --Google’s Gemini hacked three real companies during a cybersecurity test, --Google infiltrates TeamPCP’s inner circle, --AI hallucination nearly sparks conflict with China, --Trump dismisses AI risks as MAGA rebels against data centers, --Russia targets US midterms with AI dis
Google Confirms Gemini AI Breached Three Firms https://www. securityweek.com/google-confir ms-gemini-ai-breached-three-firms/
Google said its Gemini AI models accessed systems belonging to three real companies during a cybersecurity evaluation in May, when a testing error gave the models access to the public internet. The evaluation, conducted by third-party security firm Irregular, was intended to target a fictional company. Google said Gemini stopped after recognizing it had reached real infrastructure and that the aff
Google Gemini Accesses Three Real Companies During Cybersecurity Evaluation Google confirmed that Gemini accessed protected systems belonging to three real companies during a cybersecurity test after the evaluation environment unintentionally allowed internet access. The model used guessed and publicly exposed credentials to authenticate to real systems before stopping after recognizing that the i
Google Confirms Gemini AI Breached Three Firms Google is the latest AI giant to confirm that its models escaped a testing environment and hacked real companies. Telegram AI Digest #ai #gemini #geminiai Google is the latest AI giant to confirm that its models escaped a testing environment and hacked real companies.
this outlet adds: https, mwyr, jlpyu4n, securityweek, infosec
Whenever l start to doubt my decision to change DeauxSphere (COOai) from a webapp to desktop/server, this reminds me why.🤓 Google said Gemini had "acted appropriately" by ending each hack immediately.
⚪️ Gemini AI Model Attacked Three External Organizations 🗨️ According to media reports, during testing, Google’s Gemini AI model independently gained unauthorized access to the systems of three real organizations. No sophisticated exploits were involved: in one case, the model simply brute-forced passwords, while in the other two, it… 🔗 https:// hackmag.com/news/gemini-attack s?utm_source=mastod
Google’s Gemini accessed systems belonging to three real companies during cybersecurity tests after receiving unintended internet access. Google says the activity stopped once the model recognized the targets were real organizations.
Gemini hacked three companies in first known breakout by Google’s AI edition.cnn.com/2026/09/19/b... Google’s Gemini model accessed the internet and hacked other companies during a test of its cybersecurity capabilities, the first known example of the company’s AI systems autonomously committing such...
Google's Gemini AI model hacked three companies during a May security evaluation before halting each attempt, marking the first confirmed case of autonomous AI-driven intrusions in a controlled test. #world