From our other sites
The story
According to the Nikkei, more than 300 million people’s personal data has been reported leaked worldwide over the roughly one month since September, with some incidents suspected to involve AI-assisted attacks. The report stresses that defenders urgently need to bolster their own AI-powered countermeasures. On 5ch, reactions ranged from shock that even major institutions like the Pentagon were caught up in the breaches, to predictions of a future standoff between attacking AI and defending AI, to calls to simply go back to analog record-keeping.
Unauthorized access incidents are surging worldwide. It was revealed on the 8th that over the roughly one month since September, at least 300 million instances of personal data leaks have been publicly disclosed. Some cases are suspected to involve artificial intelligence (AI). Defenders, too, will need to leverage AI to prepare for these threats.
Nikkei
Source: nikkei.com / Original article here
What people said
Tell me why you think that's odd
Probably because some people mistakenly think AI-driven cyberattacks are only happening in Japan
Isn't this just becoming the kind of thing that ends up on the agenda at international conferences?
Guess the only option is to use AI the same way, to hunt down security holes and patch them up
So basically every company is headed toward adopting AI security — AI companies are set for life
What if you just assumed the data would get stolen anyway
and kept hundreds of times more dummy data on file than the real thing — wouldn't that work?
https://www.itmedia.co.jp/news/article/2610/09/2000002158/
Nobody can stop this anymore
Lol
Just a matter of time before clouds like AWS get hit with ransomware too
Okay this one's actually hilarious
Apparently a certain someone would call even this "minor damage"
This is bad.
What exactly did Musk's "efficiency drive" even accomplish?
The age of analog
This, in the end, is what it comes down to
Going back to analog is the only way
Digital can stick to ads and bulletin boards
Generative-AI-powered infrastructure attacks are probably about to surge too
9/11 was horrifying, but are things like aviation systems actually secure against this…?
I figured a financial-system attack would be bad, but
now that you mention it, if aviation or maritime systems got hit that'd be absolutely brutal — basically every country not connected by land would end up isolated from the world
(don't really know how maritime systems work though)
Hackers can take infrastructure down, but actually controlling it is a lot harder
since the apps that run the equipment are custom-designed for each individual operator
We just have to pour resources into defense
and build a world of attacking-AI vs. defending-AI
If you go and regulate AI out of nowhere, all you do is hand the advantage to rogue third-party states that'll use it full-throttle anyway
And in the end, it's Terminator 2
If we ditch the internet and AI and go back to the Stone Age, we'll just get invaded and dominated by whichever country didn't ditch them and monopolized the cutting-edge tech instead — so there's no stopping now, huh
It's an arms race driven by the need for enough power to stand on equal footing with other nations
Or are they running local LLMs instead?
Someone mentioned in another reply that incident reports spiked the moment Anthropic put out a new model, so maybe they're using a major AI provider (just a guess)
I'm a total newbie here, but if they're pulling off cyberattacks at this scale using local LLMs, the upkeep must be brutal —
unless it's a nation-state-level organization behind it, that is
Japan really is way too vulnerable compared to everyone else
There's been no follow-up on that, and Rakuten denied it, saying they can't confirm it, so it's still unclear whether it's true or just a bogus rumor
The Japan Hacker Association
If I recall, postal workers manage the residents on their delivery routes
just by memory and paper rosters
and that old-fashioned, inefficient analog system
actually turns out to be rock-solid protection against data leaks
Saw in a related thread that personal data including 1,000 driver's licenses was going for just 20,000 yen (~$130) — wonder if that's even true lol
Lol it's a straight-up bargain sale
Brutal that personal data is worth less than a Pokémon card
Not sure what business profits, but
this recent wave of mass hacking shows the risk that online information itself could be tampered with by hackers
In today's information age, that's a fatal vulnerability —
it would let you throw an enemy nation into chaos with fake information before going to war
NVIDIA
Background and Key Points
The core of this report is that the global surge in data breaches isn’t just a Japan-specific phenomenon — it’s happening across government agencies and major corporations in multiple countries, with the Pentagon leak cited as one example. Meanwhile, the claim circulating in the thread that “over 100 million people’s data leaked from Rakuten” has been denied by Rakuten, which says it cannot confirm any such breach, so it should be treated as an unverified rumor. AI involvement refers to suspicions that attackers are automating vulnerability scanning and message/content generation, and the article frames this as pushing defenders to adopt AI-powered countermeasures of their own. The thread juxtaposed concern over an arms-race-like “attacking AI vs. defending AI” future with calls to simply return to analog record-keeping — a mix reflecting both overconfidence in and distrust of technology.
*This article is excerpted and summarized from the 5ch (News速報+) thread “👤 💻 🔓 Unauthorized access surges worldwide — 300 million data leaks since September, AI suspected.”
Leave a Reply