NVIDIA research shows how agentic AI fails under attack

Help Net Security

December 8, 2025 at 7:00 AM

Overview

NVIDIA's research highlights the vulnerabilities of agentic AI systems, which operate with minimal human oversight. These systems face new risks due to their interactions with various models, tools, and data sources, necessitating a safety and security framework to address these challenges.

Key Takeaways

Affected Systems: Agentic AI systems, tools, and workflows
Action Required: Implement the safety and security framework proposed by NVIDIA and Lakera AI to assess and mitigate risks.
Timeline: Newly disclosed

Original Article Summary

Enterprises are rushing to deploy agentic systems that plan, use tools, and make decisions with less human guidance than earlier AI models. This new class of systems also brings new kinds of risk that appear in the interactions between models, tools, data sources, and memory stores. A research team from NVIDIA and Lakera AI has released a safety and security framework that tries to map these risks and measure them inside real workflows. The work … More → The post NVIDIA research shows how agentic AI fails under attack appeared first on Help Net Security.

Impact

Agentic AI systems, tools, and workflows

Exploitation Status

The exploitation status is currently unknown. Monitor vendor advisories and security bulletins for updates.

Timeline

Newly disclosed

Remediation

Implement the safety and security framework proposed by NVIDIA and Lakera AI to assess and mitigate risks.

Additional Information

This threat intelligence is aggregated from trusted cybersecurity sources. For the most up-to-date information, technical details, and official vendor guidance, please refer to the original article linked below.

Related Topics: This incident relates to NVIDIA.

Read Original Article

Related Coverage

Palo Alto Networks warns of DoS bug letting hackers disable firewalls

BleepingComputer

Palo Alto Networks has released a patch for a serious vulnerability that could enable attackers to conduct denial-of-service (DoS) attacks, effectively disabling firewall protections. This flaw allows unauthenticated individuals to exploit the vulnerability, raising concerns for organizations relying on Palo Alto's security products. The ability to disable firewalls poses significant risks, as it could lead to unauthorized access and data breaches. Companies using affected products are urged to apply the patch immediately to safeguard their networks. This incident serves as a reminder of the ongoing challenges in maintaining cybersecurity defenses against evolving threats.

Jan 15, 2026

China bans U.S. and Israeli cybersecurity software over security concerns

Security Affairs

China has ordered domestic companies to cease using cybersecurity software from over a dozen U.S. and Israeli firms due to concerns about national security. This move comes amid escalating tensions between China and the U.S. regarding technology and trade, particularly in the semiconductor sector. The ban affects various cybersecurity solutions that many Chinese companies rely on to protect their networks and data. By restricting these foreign products, China aims to bolster its own cybersecurity infrastructure and reduce dependency on foreign technology. This decision could have significant implications for international cybersecurity firms and may lead to further divisions in the global tech landscape.

Jan 15, 2026

Microsoft disrupts massive RedVDS cybercrime virtual desktop service

BleepingComputer

Microsoft has successfully disrupted RedVDS, a significant cybercrime platform that has been linked to approximately $40 million in losses reported in the U.S. since March 2025. The platform was known for offering a virtual desktop service that criminals used to conduct various illegal activities. This disruption is a major step in combating cybercrime, as it not only targets the infrastructure used by attackers but also aims to deter future criminal operations. The impact of RedVDS has been felt widely, affecting numerous victims who have suffered financial losses due to the platform's activities. By taking action against RedVDS, Microsoft is contributing to a broader effort to enhance cybersecurity and protect individuals and businesses from ongoing threats.

Jan 15, 2026

QR codes are getting colorful, fancy, and dangerous

Help Net Security

QR codes are increasingly being stylized with colors and shapes, leading to new security risks. Researchers from Deakin University have identified that these visually appealing QR codes can be manipulated in 'quishing' attacks, where attackers trick users into scanning harmful codes. The assumption that QR codes are safe can make users more vulnerable to these scams. The study proposes a new method to detect compromised QR codes based on their structural analysis. This research emphasizes the need for users to remain cautious and critically evaluate QR codes, similar to how they approach links in emails or websites.

Jan 15, 2026

Verizon outage affects over 2 million users: What 'SOS' means, refunds, more updates

Alleged Dell data compromise raises skepticism

SCM feed for Latest

A recent report has cast doubt on the authenticity of a claimed data breach involving Dell. The incident allegedly compromised a database containing over 5,000 records, which include emails from more than 2,000 employees. However, security experts and researchers are questioning the validity of this breach, suggesting the information may not be as serious as it appears. If true, this situation could expose sensitive employee data, raising concerns about privacy and security within the company. As investigations continue, it remains essential for organizations to remain vigilant about potential data breaches and to verify claims before reacting.

Jan 15, 2026