48 Vulnerabilities Uncovered In AI systems : Surge By 220%

In Cybersecurity News - Original News Source is cybersecuritynews.com by Blog Writer

Post Sharing

Since the initial disclosure of 15 vulnerabilities in November 2023, a 220% increase in vulnerabilities impacting AI systems has been discovered, bringing the total to 48 vulnerabilities.

The world’s first AI/ML bug bounty program, Protect AI, analyzes the whole OSS AI/ML supply chain for significant vulnerabilities. 

The experts discovered that specific security risks may be exploited against the supply chain tools that are used to create the machine learning models that drive AI applications.

Thousands of times a month, these open-source tools are downloaded to develop enterprise artificial intelligence systems.

The analysis highlights Remote Code Execution (RCE) as a widespread vulnerability that enables an attacker to execute commands or programs on a victim’s computer or server without requiring physical access. 

The compromised system could be fully taken over by the attacker, resulting in data breaches.

Significant Vulnerabilities In AI Systems

Remote Code Execution In PyTorch Serve:

An attacker can use this vulnerability to run arbitrary code to compromise the server hosting PyTorch Serve.

With a CVSS base score of 9.8, this vulnerability is categorized as a critical severity, and CVE is unavailable as per the maintainer’s request.

If PyTorch Serve were exposed to the network, a remote user uploading a model containing malicious code might attack it.

When the model is deployed, this code is run, which could result in remote code execution on the server.

Insecure Deserialization In BentoML

With a CVSS base score of 9.8, this vulnerability is categorized as a critical severity and is tracked as CVE-2024-2912.

This vulnerability allows remote attackers to execute arbitrary code on the server. 

An unsafe deserialization vulnerability exists in BentoML. An attacker can run any code on the server hosting the BentoML application by sending a specially crafted request.

It is recommended that you upgrade to version 1.2.5.

Regular Expression Denial Of Service (ReDoS) In FastAPI

With a High severity level and a CVSS base score of 7.5, the bug is tracked as CVE-2024-24762.
A denial of service attack may result from this vulnerability, making the server unresponsive.

FastAPI is susceptible to a ReDoS attack when parsing Form data in certain scenarios. By fully using the CPU, this vulnerability can be used to render the server unresponsive.

Server-Side Template Injection In BerriAI/Litellm

Attackers may use this vulnerability to permit the server to execute illegal commands.

In BerriAI’s litellm project, the hf_chat_template method uses the Jinja template engine to process user input without properly sanitizing it. On the server, this can be used to run arbitrary commands.

It is recommended to upgrade to version 0.109.1

The Complete List Of Vulnerabilities In AI Systems

CVE Title Severity CVSS Fixed Recommendations
CVE-2024-3025 Arbitrary file deletion / reading via path traversal in logo photo upload and download feature in anything-llm Critical 9.9 Yes Upgrade to version 1.0.0
CVE-2024-2083 Directory Traversal in /api/v1/steps in zenml Critical 9.9 Yes Upgrade to version 0.55.5
N/A per maintainer request Remote Code Execution due to Full Controlled File Write in pytorch/serve Critical 9.8 Yes Read security documentation for secure deployment.
CVE-2024-2912 RCE By Sending A Single POST Request Via Insecure Deserialization in bentoml Critical 9.8 Yes Upgrade to version 1.2.5
CVE-2024-3098 Prompt Injection leading to Arbitrary Code Execution in llama_index Critical 9.8 Yes Upgrade to version 0.10.24
CVE-2024-2221 Remote Code Execution via Arbitrary File Overwrite Using Path Traversal in qdrant Critical 9.8 Yes Upgrade to version 1.8.0
CVE-2024-1520 OS Command Injection in lollms-webui Critical 9.8 Yes Upgrade to version 9.1
CVE-2024-2029 Command injection in audioToWav in mudler/localai in localai Critical 9.8 Yes Upgrade to version 2.10.0
CVE-2024-3271 safe_eval bypass lead to RCE (Command Injection) in llama_index Critical 9.8 Yes Upgrade to version 10.26
CVE-2024-1600 Local File Inclusion in lollms-webui Critical 9.3 Yes Upgrade to version 9.5
CVE-2024-3573 Local File Read (LFI) due to scheme confusion in mlflow Critical 9.3 Yes Upgrade to version 2.10.0
CVE-2024-1643 join any organization and read/modify all data in lunary Critical 9.1 Yes Upgrade to version 1.2.2
CVE-2024-1740 removed user from a org can read/create/modify/delete logs in lunary Critical 9.1 Yes Upgrade to version 1.2.7
CVE-2024-1626 idor bug to change any org project in lunary Critical 9.1 Yes Upgrade to version 1.0.0
CVE-2024-0404 Mass assignment in account creation from invitation in anything-llm Critical 9.1 Yes Upgrade to version 1.0.0
CVE-2024-3029 Deactivate Multi-User Mode and Delete All Users in anything-llm Critical 9.0 Yes Upgrade to version 1.0.0
CVE-2024-1522 Remote Code Execution Via Cross-Site Request Forgery in lollms-webui High 8.8 Yes Upgrade to version 9.2
CVE-2024-1540 [gradio-app/gradio] Secrets exfiltration via the [deploy+test-visual.yml] workflow in gradio High 8.6 Yes Upgrade to commit d56bb28df80d8db1f33e4acf4f6b2c4f87cb8b28
CVE-2024-1646 Insufficient protection over sensitive endpoints in lollms-webui High 8.2 Yes Upgrade to version 9.3
CVE-2024-25723 Improper Access Control leads to Account Takeover/Privilege Escalation in zenml High 8.1 Yes Upgrade to version 0.56.2
CVE-2024-0798 privilege escalation bug to delete the uploaded document in anything-llm High 8.1 Yes Upgrade to version 1.0.0
CVE-2024-0549 Path traversal leads to anythingllm.db deletion in anything-llm High 8.1 Yes Upgrade to version 1.0.0
CVE-2024-24762 Content-Type Header ReDoS in fastapi High 7.5 Yes Upgrade to version 0.109.1
CVE-2024-3569 DOS attack in Just me mode in anything-llm High 7.5 Yes Upgrade to version 1.0.0
CVE-2024-1625 idor bug to delete any org project in lunary High 7.5 Yes Upgrade to version 1.0.1
CVE-2024-1728 Local File Inclusion in gradio High 7.5 Yes Upgrade to version 4.19.2
CVE-2024-2217 Unauthorized access to config.json file in chuanhuchatgpt High 7.5 Yes Upgrade to version 20240310
CVE-2024-1892 Denial of Service when parsing downloaded XML content in XMLFeedSpider in scrapy High 7.5 Yes Upgrade to version 2.11
CVE-2024-1739 creating account with same email in lunary High 7.5 Yes Upgrade to version 1.0.2
CVE-2024-1601 SQL injection in delete_discussion()in lollms-webui High 7.5 Yes Upgrade to version 9.2
CVE-2024-1561 Local file read by calling arbitrary methods of Components class in gradio High 7.5 Yes Upgrade to version 4.13.0
N/A per maintainer request Bypass private/linklocal/loopback IP validation Method lead to SSRF in netaddr High 7.5 Yes Upgrade to version 0.10.0
CVE-2024-3572 Parsing XML content using insecure function in scrapy High 7.5 Yes Upgrade to version 2.11.1
CVE-2024-3574 Authorization header leaked to third party site and it allow to hijack victim account in scrapy High 7.5 Yes Upgrade to version 2.11.1
CVE-2024-2206 Insufficient SSRF protection allow gradio app to proxy arbitrary URLs in gradio High 7.3 Yes Upgrade to version 4.18
CVE-2024-3283 Mass assignment that leads to privilege escalation attack in anything-llm High 7.2 Yes Upgrade to version 1.0.0
CVE-2024-3028 User can read and delete arbitrary files in anything-llm High 7.2 Yes Upgrade to version 1.0.0
CVE-2024-3101 Users can escalate privileges by deactivating ‘Multi-User Mode’. in anything-llm Medium 6.7 Yes Upgrade to version 1.0.0
CVE-2023-6568 Reflected POST XSS in mlflow Medium 6.5 Yes Upgrade to version 2.9.0
CVE-2024-3571 Local File Inclusion (LFI) to Remote Code Execution in langchain Medium 6.5 Yes Upgrade to version 0.0.353
CVE-2024-1183 ssrf bug to scan internet network in gradio Medium 6.5 Yes Upgrade to version 4.11
CVE-2024-1455 Billion laughs vulnerability that leads to DOS in langchain Medium 5.9 Yes Upgrade to version 0.1.35
CVE-2024-1729 timing attack to guess the password in gradio Medium 5.9 Yes Upgrade to version 4.19.2
CVE-2024-1599 bypass payment and create more project than limit without paying extra money in lunary Medium 5.3 Yes Upgrade to version 1.0.0
CVE-2024-1569 Denial of Service in lollms-webui Medium 5.3 Yes Upgrade to version 9.2
CVE-2024-1727 CSRF allows attacker to upload many large files to victim in gradio Medium 4.3 Yes Upgrade to version 4.19.2
CVE-2024-2260 Session fixation lead to bypass authentication in zenml Medium 4.2 Yes Upgrade to version 0.56.2
CVE-2024-3568 Transformers has a Deserialization of Untrusted Data vulnerability in transformers Low 3.4 Yes Upgrade to version 4.38

Hence, this pro-active method of detecting and resolving security issues in AI systems gives everyone significant information about vulnerabilities and facilitates their prompt fix to these vulnerabilities. 

The post 48 Vulnerabilities Uncovered In AI systems : Surge By 220% appeared first on Cyber Security News.