Sep 28, 2026 · 6 min · 3 segments
Sort of. A novel attack against RSA algorithms that forges keys rather than finds the prime factors could be a cause for concern, especially with AI enhanced computing capabilities, before we've even…
Chloe TonisHost
Before, it was thought that the only way to crack RSA was through factoring, because the way that the key was generated was by multiplying very large prime factors together.
The idea is that it's easy to do that operation, multiplying two numbers together, if you know what they are.
It's hard to take a very large number, find all of the factors of it, and find specifically the prime factors of it.
You could still crack this in theory, but with most computational capabilities, it would take millions of years.
Although one thing that's important to know is that RSA is not immune to a potential quantum cryptography capability.
We have not hit mainstream quantum computing quite yet, but a new method was discovered that potentially makes it easier to crack RSA even before we hit that, especially for some institutions that already have a lot of computing behind them.
Some researchers discovered that you don't actually need to factor these keys at all.
You can perform a key forgery type of attack, one that is actually completely practical for 1024-bit RSA keys, and even potentially for 2048 and 4096-based keys.
Many security standards, including NIST, should provide a level no less than 128 bits.
This attack effectively reduces the levels for 1024 to 65 bits, 2048 to 90 bits, and 4096 to 119 bits.
And these numbers may actually be even smaller due to the AI stuff I mentioned, if we decide to actually leverage that in the future.
It could actually bring those numbers down quite a bit and further undermine the security of the algorithms, especially if this attack is practical and holds up to peer review.
I will admit the attack is pretty technical and I don't quite understand all the math that goes into it, but essentially what happens is that no one ever actually finds the prime factors that make up the RSA key.
but it does do the problem more efficiently by incorporating part of the mathematical work with the responses that are received from the RSA signing oracle.
This attack essentially is able to abuse legitimate signing functions to obtain results they're not really supposed to obtain, and use those results to find out or forge a key a lot faster.
This algorithm isn't exactly new, it has actually been around since 2007, but new work around it has made it potentially possible to actually break RSA encryption, though it's not quite a universal break.
Even though we haven't successfully universally broken RSA, it is important to note that this is technology that does currently exist and that some people do have access to.
mostly educational institutions or state-sponsored attackers that have a lot of resources behind them due to some government backing.
Before, it was thought that the only way to crack RSA was through factoring, because the way that the key was generated was by multiplying very large prime factors together.
The idea is that it's easy to do that operation, multiplying two numbers together, if you know what they are.
It's hard to take a very large number, find all of the factors of it, and find specifically the prime factors of it.
You could still crack this in theory, but with most computational capabilities, it would take millions of years.
Although one thing that's important to know is that RSA is not immune to a potential quantum cryptography capability.
We have not hit mainstream quantum computing quite yet, but a new method was discovered that potentially makes it easier to crack RSA even before we hit that, especially for some institutions that already have a lot of computing behind them.
Some researchers discovered that you don't actually need to factor these keys at all.
You can perform a key forgery type of attack, one that is actually completely practical for 1024-bit RSA keys, and even potentially for 2048 and 4096-based keys.
Many security standards, including NIST, should provide a level no less than 128 bits.
This attack effectively reduces the levels for 1024 to 65 bits, 2048 to 90 bits, and 4096 to 119 bits.
And these numbers may actually be even smaller due to the AI stuff I mentioned, if we decide to actually leverage that in the future.
It could actually bring those numbers down quite a bit and further undermine the security of the algorithms, especially if this attack is practical and holds up to peer review.
I will admit the attack is pretty technical and I don't quite understand all the math that goes into it, but essentially what happens is that no one ever actually finds the prime factors that make up the RSA key.
but it does do the problem more efficiently by incorporating part of the mathematical work with the responses that are received from the RSA signing oracle.
This attack essentially is able to abuse legitimate signing functions to obtain results they're not really supposed to obtain, and use those results to find out or forge a key a lot faster.
This algorithm isn't exactly new, it has actually been around since 2007, but new work around it has made it potentially possible to actually break RSA encryption, though it's not quite a universal break.
Even though we haven't successfully universally broken RSA, it is important to note that this is technology that does currently exist and that some people do have access to.
mostly educational institutions or state-sponsored attackers that have a lot of resources behind them due to some government backing.
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.
