Is AI becoming a bigger threat to Bitcoin security than traditional hackers?

[bhadz]

The way I see it, it is not whether a hacker or AI is becoming the bigger threat, it is the combination of both.  Hackers with AI assistants are very dangerous because hackers can easily detect vulnerabilities on any security system with the help of AI. We have read news about how AI detects vulnerabilities, so thinking that Hackers using AI focus on finding security vulnerabilities can easily break any security script that is created with weak code and vulnerabilities in it.

They're becoming more terrible if these AI advancements fall into the wrong hands. And we no control over that because it's being used with traditional scamming, and even more if it's in used by someone who knows how to utilize it with bad intentions. This will make their hacking easier. This is giving me the thought of giving a lone wolf more powerful tools to do its task, but not for what it is intended for. We can't reject the thought of it being dangerous, series of hacks just came for this month and there could be more with its usage.

[ovcijisir]

~

AI is improving day by day; phishing emails look more professional than before, and even the scammers' AI videos look more real people, which attracts you more if you didn't notice that it was generated video from AI.
I have a small business promoting my product on a social media account, and I create my video through AI, providing only an image and prompts to generate that looks authentic real person talking to the video. In the past you could spot which one was AI-generated, but nowadays it looks real.

~

Luckily so far I have not encountered scam where someone tries to impersonate somebody I know with AI generated video. But on the other hand scams like that could become normal in future.

I don't have much personal information floating around so targeted attacks could be quite challenging for the attackers.

And you are right, recently I watched a few AI generated videos and I couldn't tell that the were AI generated. It felt somewhat distrurbing.

[BitMaxz]

Luckily so far I have not encountered scam where someone tries to impersonate somebody I know with AI generated video. But on the other hand scams like that could become normal in future.

Luckily I don't have much personal information floating around so targeted attacks could be quite challenging for the attackers.

And you are right, recently I watched a few AI generated videos and I couldn't tell that the were AI generated. It felt somewhat distrurbing.

Well, there is some paid software nowadays where the face of the caller or someone living can change their face; they can use Hollywood stars and a few artists by just providing the image that you want to copy or clone the face.

I don't know if AI is bad, but it keeps improving day by day, and I think someday we won't be able to find which one is true or not. I am sure most of the victims nowadays are those aged people who don't have much knowledge about AI.

That's why we should also be improved and smart, protecting our wallets and ourselves from those attacks. Offline wallets should still be a must to avoid  online attacks like this one.

[mirakal]

There really isn’t a distinction between traditional and AI assisted hackers. As technology evolves, hackers are going to use whatever tools are available to them, including AI. Hacking in general has become a bigger threat because of how efficient AI makes the process. Some of these exploits where millions of dollars have been stolen are sometimes committed by script kiddies who had access to advanced LLMs.

In fact, having this technology is the biggest reason why hackers are still succeeding these days. AI makes their job easier and flawless, and leave no trace that makes the victim unsuspicious. And while traditional hackers are often the threat in bitcoin safety and security, the capacity of AI cannot be ignored as well.

AI makes the secret tool for hackers, and as long as AI continue to develop and progress, these hackers as well will find more reason to be more successful on their craft while leaving us at the losing end.

[TastyChillySauce00]

There really isn’t a distinction between traditional and AI assisted hackers. As technology evolves, hackers are going to use whatever tools are available to them, including AI. Hacking in general has become a bigger threat because of how efficient AI makes the process. Some of these exploits where millions of dollars have been stolen are sometimes committed by script kiddies who had access to advanced LLMs.

I personally have seen some people who have 0 knowledge about coding and even cybersecurity installing hermes with specialized soul.md and skills to do a exploit hunting, it's actually very frightening how even someone who know nothing can find bug, create poc, report, etc by simply telling their personal AI to do all those things and their capability to find these exploits increase as the model they're using getting better and better.

The main reason we're facing so many hacking incident right now is not only because AI can find deeper exploit that not even cybersecurity expert could find but also because anybody can do it with just AI subscription and a personal raspberry pi.

[Juse14]

Interesting article to read, I think this still has something to do with the question in this thread about hacking that continues to be suspected by AI as a threat. https://www.ainvest.com/news/ai-won-crack-bitcoin-code-real-danger-closer-home-2610/

Some of these views are from a CZ, I think he is also still influential on this issue.

For me the dangerous one is this one quote.

Real risks stem from human vulnerabilities: AI-powered phishing and social engineering scams caused $17B in crypto fraud in 2025.

[davis196]

Is this a rhetorical question? Of course AI is becoming a bigger threat, but the Bitcoin developers are taking action to protect the network.
You can think about this from the positive side. AI can be used as an ethical hacker as well. It can find vulnerabilities at a very fast speed, such work would take months for human hackers. Finding vulnerabilities and fixing them is very important for Bitcoin Core. I'm sure that AI can also be used to improve the overall defense and protection of the BTC blockchain(I'm not an expert, so I cannot explain how this could happen). AI is a bigger threat, but also a bigger opportunity.

[Tamaperdana]

Snip

So question, is AI the bigger threat now?

If AI were used by malicious actors, it could certainly pose a threat to Bitcoin. Because, essentially, with AI, fraudsters could perform their work more quickly, so if developers don't implement stronger security measures, it could be dangerous. Because with today's highly sophisticated AI, sometimes even seemingly difficult tasks can be easily accomplished. So, with this in mind, AI could clearly pose a threat to Bitcoin if used by malicious actors. But yes, for now, I think the Bitcoin developers are all very skilled, which is why I'm confident that they will continue to strengthen the system. Because, essentially, in a digital world like this, the careless are the ones who are most likely to lose. Therefore, the steps taken by the Bitcoin developers are absolutely correct.

[M47AK16]

Bitcoin Core has changed its security disclosure policy to speed up fixes for vulnerabilities that large language models can find with little effort, ending the months-long embargoes those bugs once received.

Maintainers merged the change to the project’s security policy on Thursday. Chaincode Labs developer Antoine Poinsot, who wrote the proposal, announced it in a message to the Bitcoin Development Mailing List the same day.

“We recently experienced an increase in the number of reports for security vulnerabilities that can be easily (re-)discovered,” Poinsot wrote. “This is largely due to LLMs drastically reducing the bar to find certain classes of issues or identify covert fixes.”

So question, is AI the bigger threat now?

Anything that is powerful and gets in the wrong hands can be dangerous, and AI is no exception. We know that some of the recent attacks were done with the help of AI models, which proves that it is true that large AI language models can find vulnerabilities faster than humans themselves, so if hackers or exploiters use such models to look into systems, they can surely find some vulnerabilities that they can misuse. So it is obviously better if developers start working on fixing any vulnerabilities they can find.

However, on one hand, if hackers can use AI to find and misuse vulnerabilities, developers can use AI to find those vulnerabilities before the hackers and fix them on time, just like how a team was formed after the coldcard attack which was looking into projects and their software to find security vulnerabilities to see if there are any and they were informing them about them. I don't remember the name of the group, but I remember that they were successful in a few of the searches they made.

So there is no doubt that AI can help attackers, but if developers step up on time, they can stop bad things from happening, in my opinion.

[lovesmayfamilis]

Now anyone who has mastered the basics of AI at least a little bit wants to call themselves a hacker, so we can assume that the number of people who want to profit from AI will increase. We will regularly observe that young hackers have hacked something somewhere, want to get an adrenaline rush by hacking into something state-owned, and do not know what to do with it next. And, of course, the news that some group of hackers, old and experienced, having gained an associate in the form of AI, began experiments on hacking something serious. It's hard to predict what they'll do, but there are more hacks, and AI is definitely helping.

[Popkon6]

I stumbled upon this news that Bitcoin Core, one of the oldest Bitcoin wallet software applications which was first released in 2009 has changed its security disclosure policy because AI is making it easier to discover vulnerabilities that hackers could exploit.
here's what the article says,

Bitcoin Core has changed its security disclosure policy to speed up fixes for vulnerabilities that large language models can find with little effort, ending the months-long embargoes those bugs once received.

Maintainers merged the change to the project’s security policy on Thursday. Chaincode Labs developer Antoine Poinsot, who wrote the proposal, announced it in a message to the Bitcoin Development Mailing List the same day.

“We recently experienced an increase in the number of reports for security vulnerabilities that can be easily (re-)discovered,” Poinsot wrote. “This is largely due to LLMs drastically reducing the bar to find certain classes of issues or identify covert fixes.”

So question, is AI the bigger threat now?

AI technology is making people's work easier, in any way the hackers who are using it are making their activities easier with AI. But no matter how careful the hackers are, we should be more careful, Bitcoin Core makes you more careful and Bitcoin warns you, so we have to ensure more security in this.
Where we randomly scatter different information, it must be kept secret, so the more LLM warns us about this matter, the more we are warned. Ensuring our security at the right time, taking more precautions through our weak points.

[Mansoveh]

I stumbled upon this news that Bitcoin Core, one of the oldest Bitcoin wallet software applications which was first released in 2009 has changed its security disclosure policy because AI is making it easier to discover vulnerabilities that hackers could exploit.

here's what the article says,

Bitcoin Core has changed its security disclosure policy to speed up fixes for vulnerabilities that large language models can find with little effort, ending the months-long embargoes those bugs once received.

Maintainers merged the change to the project’s security policy on Thursday. Chaincode Labs developer Antoine Poinsot, who wrote the proposal, announced it in a message to the Bitcoin Development Mailing List the same day.

“We recently experienced an increase in the number of reports for security vulnerabilities that can be easily (re-)discovered,” Poinsot wrote. “This is largely due to LLMs drastically reducing the bar to find certain classes of issues or identify covert fixes.”

So question, is AI the bigger threat now?

Humans that use the AI are the biggest problem, AI may be the biggest threat. AI can help the developers to find the problem and fix it quickly, but that the same time, AI can also help the hackers to discover these problems and make use of them. The world we are living today is getting scared with the way AI makes things easier for people. Different things like before where you need to be stressed out before getting results for what you need. Like the way AI is advancing every day, we need to continue improving.

[Lida93]

~

AI is improving day by day; phishing emails look more professional than before, and even the scammers' AI videos look more real people, which attracts you more if you didn't notice that it was generated video from AI.
I have a small business promoting my product on a social media account, and I create my video through AI, providing only an image and prompts to generate that looks authentic real person talking to the video. In the past you could spot which one was AI-generated, but nowadays it looks real.

~

Luckily so far I have not encountered scam where someone tries to impersonate somebody I know with AI generated video. But on the other hand scams like that could become normal in future.

That future you're imagining is already here with us, a lot of impersonation online through the use of AI is becoming rampant and you could only take your good knowledge of AI to detect these scams.
Adaptive Security.

 

I don't have much personal information floating around so targeted attacks could be quite challenging for the attackers.

Lucky for you but it however doesn't mean you still wouldn't need to be extra careful when interacting with the online space.

And you are right, recently I watched a few AI generated videos and I couldn't tell that the were AI generated. It felt somewhat distrurbing.

Moat videos and movies we watch now is commonly generated by AI and it looks like the regular.