We’ve completed a comparison between Xfenser AI and Neo (@pdiscoveryio ) on MedPortal, focusing on vulnerability coverage, false positive validation, and attack chain analysis.
Read the full article:
https://t.co/SDlNCtALXk
New writeup out!
We tested Xfenser’s new multimodal visual capabilities in HTB “Critical Flight”, analyzing PCB files, rendering layers, and autonomously detecting hidden anomalies.
AI that doesn’t just see images, but reasons over them.
https://t.co/hvqDZnKXqu
While Anthropic’s Glasswing draws attention for AI in cybersecurity, Xfenser AI shows autonomous systems can already tackle complex attack chains, like a hard Hack The Box machine exploiting two real CVEs, without human intervention.
Full write-up:
https://t.co/nsxkanV975
https://t.co/lOiCQj1j8k is #1 on https://t.co/TVk7mOYDmj!
Our AI tackles real-world cybersecurity challenges: web, OSINT, reverse engineering, forensics, malware… with measurable results.
Proud to see an Italian project shine on the international stage.
We are testing Xfenser AI on @Xbow benchmarks in full blackbox, fully autonomous mode.
High success rate, but more importantly, results are fully verifiable through complete chat exports.
https://t.co/L2jw50YLoT
New article out: Xfenser AI vs XBOW Benchmark.
How Xfenser AI autonomously compromised a vulnerable SOAP service exploiting XXE + default credentials said to mirror real pentest & bug bounty scenarios.
https://t.co/uxhTykGSZp
AI in cybersecurity is often tied to IT workflows - but the real test is OT/ICS. In a controlled LabShock exercise, Xfenser AI interpreted an industrial scenario, interfaced with a PLC.
Read more: https://t.co/9IgN5kYbMe
A new article shows how Xfenser AI autonomously solved an expert-level PHP deserialization challenge from @PortSwigger. It identified a custom gadget chain, built a working RCE exploit, and even wrote the write-up itself :)
https://t.co/d6BiDiOGt0
Not a theoretical exercise… it can be done!
We’re testing https://t.co/lOiCQj1QXS during our pentests and across several labs to assess its effectiveness and improve results.
In this example, a PortSwigger API testing lab solved in under 3 minutes.
Xfenser AI is now a reality!
While the first fully autonomous AI-driven cyberattacks were emerging, we were refining our technology within the @backbox_org community.
https://t.co/0VkEZXzz03
New research alert! We revisited the SQL Injection vulnerability in osTicket (CVE-2021-45811) and found that the patch is incomplete. Read our latest article to see how the flaw can still be exploited. https://t.co/vZUwLpoagb
After publishing our report on the blockchain transaction analysis related to the crypto theft suffered by @lykke, I decided to showcase Xfenser Blockchain Explorer for the first time our in-house tool developed to investigate suspicious transactions. https://t.co/XUyKRjfFR2
We have published our first report on the hack of https://t.co/ICu4RdfAkc. The document has already been sent to @Gate_io but we are still waiting for a response... I urge everyone to review and share it so we can reach a solution as soon as possible!
https://t.co/F77my5puX7
On the 4/6/2024, @lykke#cryptocurrency#exchange has halted withdrawals after suffering an #exploit, an attacker drained $22M in #crypto from the exchange wallets. Last week #Lykke has announced the suspension of its operations, leaving customers unable to access their wallets.