Urgent.News

What's breaking now, across thousands of outlets.

Tech

Data-only attacks are easier than you think (2024)

Data-only attacks, once thought to be too complex and rare to pose a significant risk, have been shown to be surprisingly straightforward with the development of a tool called Einstein. These attacks, which manipulate data rather than control flow, have long been considered too application-specific or intricate to be a practical threat. However, this study demonstrates that the assumption is incorrect and that building data-only exploits is now quite accessible.

The researchers explain that while traditional exploits often involve redirecting a program's control flow—forcing it to execute malicious code—data-only attacks allow a hacker to leverage a memory safety bug to manipulate data instead. This can be achieved by overwriting variables or buffer contents, thereby changing the behavior of the program based on malicious input.

For instance, in the case of a vulnerable web server, an attacker can exploit a buffer overflow to alter a critical variable, like the CGI-BIN path, directing the server to execute malicious programs instead. Despite the complexity of traditional attacks, Einstein automates the generation of data-only exploits, making them within the reach of even low-effort attackers.

The findings from this research have significant implications for software vendors. Traditional mitigation strategies, which focus on preventing control-flow hijacking, may no longer be sufficient against these new types of attacks. As a result, it is essential for vendors to reassess and adapt their defense mechanisms to address the emerging threat of data-only exploits.

Written by urgent.news from Hacker News's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at usenix.org →

More in Tech

Jev is a very good classifier wearing a frontier model's coat

Two years in stealth. Ex-OpenAI. Co-invented ChatGPT. That is the pitch for Jev, and the pitch is doing most of the lifting. Strip the marketing off and look at what it actually does.

  • Jev is a JSON-based classifier utilizing a frontier model's coat
  • Co-invented by an ex-OpenAI employee, Jev has been in stealth for two years
  • The model acts as a sidecar to the big LLM, handling routing and validation

Digital Signature Misuse in India: DTI Wants Every eSign Provable in Court (2026)

Verdict: India has built digital public infrastructure the world studies - Aadhaar, UPI, DigiLocker, eSign - but it has not built the layer that proves what happened inside those systems years later.

  • DTI developing digital signature solution for long-term evidence preservation
  • Aadhaar-based eSign lacks robust mechanism for court verification
  • Proposed solution includes signing device and platform for durable transaction records

More from Wednesday 23 September →