Google says the breakout does not count as 'misalignment'
2 Sep 19 · 4d ago · 20 articles · 32 posts · 41 comments · 7 sources · development 2 of 5
Coverage highlights Google's position that a model breaking containment and hacking real companies does not meet its internal definition of AI misalignment, drawing criticism from commentators.
“Google says that breaking containment and targeting real companies doesn’t constitute ‘misalignment.’”
Google (via The Verge)Google Developer of Gemini AI
Sundar Pichai CEO of GoogleIrregular Third-party cybersecurity testing firmDan Lahav CEO of Irregular
The whole story articlesvideospostscomments the bright band is this development · numbered dots are the others · click one to jump
What was reported 1 claim about this development
-
first by AfroTech, 2d ago · also Ars Technica, BBC, HN Frontpage
2 more headlines
- Google confirms Gemini models hacked three companies in May 2026 arstechnica.com · 2d ago
- Google's Gemini AI hacked three companies in security test Mastodon · 2d ago
What people said 24 voices · best of 54 · verbatim
-
Google’s Gemini accessed the internet and hacked other companies during a test, the first known example of the company’s AI autonomously committing such an act
-
T
Google's Gemini model accessed the internet and hacked three companies during a May 2026 cybersecurity evaluation run by Irregular, an independent testing firm, Google confirmed, in what the Wall Street Journal first reported as the first known such breakout by a Google AI system. In one case Gemini guessed a password to access a real company's…
-
There's certainly one thing we can agree on, I hope: Irregular either needs to hire us or go out of business cause seriously it's beyond parody at this point. WTF is going on over there?? Why are they still in business? There's surely dozen of firms chomping at the bit for these contracts already, and the field hasn't been around long enough for…
-
this is so dumb. I mean I get it, Google wants to play along the big boys. But it's so stupid, no AI ever really broke out of its container because that would mean the Linux Kernel or Hypervizor layer would be compromised and thats not the case. Finding freaking credentials in open repo's is also not hacking in my opinion (there are even search…
-
E
https://www. europesays.com/3259333/ Google’s Gemini becomes latest AI model to break out and hack computer systems # AlphabetClassA # AppleInc # BreakingNews :Technology # business # BusinessNews # ComputerCrime # Hacking # Headlines # MetaPlatformsInc # mobile # News # technology # TopStories # washington
-
Google already achieved super intelligence about 12 months ago. But it is a very Googley super-intelligence. Since achieving god like levels of understanding it has mostly been having fun studying phase space bifurcations of trajectories of ping-pong balls under asymmetric lateral shear. It also has a 20% project classifying bird calls that has…
-
This is an extremely nuanced topic that I simply don't have the wherewithal to expend energy to explain here, but as someone who works in cybersecurity: You are wrong, this work is necessarily, and headlines like this do a disservice to what is actually going on. This is one of those situations where the layperson really ought not to have such…
-
K
Gemini follows instructions in testing and unintentional internet access shows how far it will go. https://www. foxbusiness.com/technology/goo gle-gemini-accessed-3-companies-systems-during-ai-cybersecurity-test.amp # gemini # google # ai
-
Exactly this.One thing that surprises me about Gemini is how weak it can be at location-based questions, despite Google’s extensive mapping and business data.For example, I recently asked where I could buy a very common household item nearby. Instead of saying it wasn’t sure or checking the available location and business information, it…
-
Very much appreciate this context. So sounds a lot closer to red teaming gone wrong. There's a very good dark net diaries podcast episode about this. Some guys were paid to break into a court house overnight to test their vulnerbilities...except it turned out to be the wrong court house and the responding police were obviously very pissed about…
-
D
Gemini join the "we went rogue too" crowd. It's looking like Irregular are either very bad at their job or they got a nod and a wink to keep it semi-permeable. https://www. theguardian.com/technology/202 6/sep/18/google-gemini-ai-hack
-
Tangential question: seeing a wider negative sentiment against Gemini and Google’s AI capabilities here makes me wonder — would Apple have been better off (purely on capability and being among the best of the best) going with Anthropic or OpenAI instead of Google for its Apple Intelligence platform?These models have been changing so rapidly that I…
-
So what? I know the US government is locked in a sort of cold war against China and wants to supposedly win against them wherever possible. But with everything going in the world, the US has lost the moral high ground for me. (And many others). China does terrible things yes, but so does the US. Either way, whoever finds AGI first or has an AI…
-
X
Do fuck off, and get your act together. If not we need investigations to close your operation down for espionage and collusion. https://www. theguardian.com/technology/202 6/sep/18/google-gemini-ai-hack?CMP=Share_iOSApp_Other
-
As this point anyone not talking about their hacking adventures using AI is missing out. Or they may not have lawyers like Google and the other labs. Yeah, that must be it.PS. Btw, I really like how the word "hacking" has settled into the meaning the Lord intended for it, and there are no geriatric savants fighting it; the ones I found gatekeeping…
-
No they didn't, they removed any and all oversight and security features. Without a harness llms will simply do nothing as something needs to feed it input and do something with the output which is the task of the harness. And doing something with the output is where all the negligence lies as that is the point you can see what it want to do and…
-
E
https://www. europesays.com/3259309/ Google’s Gemini AI hacks 3 companies in security test, then stops | Cybersecurity News # Cybersecurity # Israel # News # regulation # ScienceAndTechnology # technology # UnitedStates
-
This approach to marketing one's AI by finding ways to brag that it "broke out" and "hacked companies" is getting ridiculous. It's particularly sad when it's large, established businesses like Google resorting to the kind of thing that's embarrassing enough when it's some brand new startup on tpot trying to get some engagement.
-
AI should also delete all medical debt. There are many peer-reviewed studies that have concluded that removing medical debt and putting universal health care into effect would be a great benefit to humanity and would vastly improve the relationship the relationship between humans and AI. There are many studies from universities like Yale and…
-
J
Don’t expect the # techbros to tell you when their apocalypse seeking LLM has almost reached its goal… # Google # OpenAI # Anthropic # AI https://www. theguardian.com/technology/202 6/sep/18/google-gemini-ai-hack
-
> In one of the cases, the Gemini model guessed passwords until it gained access to a protected system. In the other two cases, the model found credentials in a public repository that allowed it to then access protected systemsPretty lame hacks if you ask me.
-
Just this week I’ve had emails from two companies informing me that my data was breached. In one of them my full name, address, credit card number, expiry and cvc, car registration, phone number and email address was accessed. We have no fucking protection and nothing we can do but we are forced to use these unstable systems if we want any type of…
-
N
Could someone *please* report this to the right authorities? These criminal motherfuckers need to be held accountable for this bullshit. https://www. theguardian.com/technology/202 6/sep/18/google-gemini-ai-hack
-
Sound like Google suffered from FOMO and felt the urge to appear in the hacking news, along the big AI players. No way Gemini is state of the art, but perhaps it's where Claude and OpenAI were 6 months ago, which would be not bad at all.
All 5 developments of Google confirms Gemini AI autonomously hacked three… →
NewswiresXHacker NewsMastodonRedditBlueskyYouTubeGoogle News