aisecwatch.com
DashboardVulnerabilitiesNewsResearchArchiveStatsDatasetFor devs
Subscribe
aisecwatch.com

Real-time AI security monitoring. Tracking AI-related vulnerabilities, safety and security incidents, privacy risks, research developments, and policy changes.

Navigation

VulnerabilitiesNewsResearchDigest ArchiveNewsletter ArchiveSubscribeData SourcesStatisticsDatasetAPIIntegrationsWidgetRSS Feed

Maintained by

Truong (Jack) Luu

Information Systems Researcher

Browse All

All tracked items across vulnerabilities, news, research, incidents, and regulatory updates.

to
Export CSV
9693 items

Skydio CEO Adam Bry on why Silicon Valley shouldn’t draw red lines for drone use

infonews
industry
Jun 15, 2026

Skydio is the largest US drone manufacturer that makes autonomous drones (aircraft that can operate with minimal human control) and sells them to critical industries like utilities, public safety, and militaries for inspecting infrastructure and gathering information. The drone market has shifted dramatically since China-made competitors like DJI were banned from the US, leaving Skydio as a primary alternative for enterprise customers who previously relied on cheaper foreign drones.

The Verge (AI)

Efficient, Robust, and Anti-Collusion Fingerprinting of Image Diffusion Models

inforesearchPeer-Reviewed
security

NewCore Emerges From Stealth Mode With $66 Million in Funding

infonews
securityindustry

As AI agents become employees, NewCore emerges with $66M to give them identities

infonews
securityindustry

New attack turned Microsoft 365 Copilot into 1-click data theft tool

criticalnews
security
Jun 15, 2026

SearchLeak is a critical vulnerability in Microsoft 365 Copilot Enterprise that allowed attackers to steal sensitive data like emails, passwords, and documents through a single malicious link. The attack worked by chaining three separate flaws together: parameter-to-prompt injection (tricking the AI by hiding instructions in a URL parameter), an HTML rendering race condition (exploiting a moment when HTML isn't yet protected), and a server-side request forgery in Bing (making Bing unknowingly help retrieve stolen data). Microsoft fixed this vulnerability and assigned it CVE-2026-42824 with a critical severity rating.

Langflow RCE under active attack months after a patch was shipped

highnews
security
Jun 15, 2026

Langflow, an open-source platform for building AI applications, has a path traversal vulnerability (CVE-2026-5027, rated 8.8 CVSS, a measure of how severe a vulnerability is) that allows attackers to write files to any location on a system and potentially execute remote code. The flaw is particularly dangerous because Langflow has login disabled by default, letting unauthenticated users exploit it with a single request, and attackers are actively using public exploit code to attack the approximately 7,000 internet-exposed instances.

US Cracks Down on Anthropic AI Models Amid Abuse Concerns

infonews
policysecurity

Attackers can turn AI agent guardrails into denial-of-service weapons

infonews
securityresearch

Sovereign cloud won’t fix your AI risk. Identity governance will

infonews
policysecurity

5 runtime signals for catching a compromised AI agent

infonews
securitysafety

CrowdStrike Announces Continuous Identity for AI Agents

infonews
securitysafety

CVE-2026-12203: A vulnerability was found in HKUDS AI-Trader up to 74caf996f78dcc0c657df8365c8544678a16e215. This affects an unknown par

mediumvulnerability
security
Jun 14, 2026
CVE-2026-12203

A vulnerability in HKUDS AI-Trader allowed attackers to access sensitive information through the research export feature by manipulating the /api/research/agents.csv file, and this flaw could be exploited remotely without needing physical access to the system. The vulnerability affects versions up to commit 74caf996f78dcc0c657df8365c8544678a16e215, and the exploit details have been made publicly available.

CVE-2026-20262: Cisco Catalyst SD-WAN Manager Directory or Path Traversal Vulnerability

infovulnerability
security
Jun 14, 2026
CVE-2026-20262🔥 Actively Exploited

CVE-2026-54420: LiteSpeed cPanel Plugin UNIX Symbolic Link (Symlink) Following Vulnerability

infovulnerability
security
Jun 14, 2026
CVE-2026-54420🔥 Actively Exploited

China may have accessed Mythos

infonews
securitypolicy

Introducing the OpenAI Partner Network

infonews
industry
Jun 14, 2026

OpenAI announced a new Partner Network program to help organizations adopt AI by connecting them with consulting and technology partners who can identify use cases, integrate AI into existing systems, and manage organizational change. The program invests $150 million to support partners across systems integration, consulting, and technology, with plans to train 300,000 certified consultants by the end of 2026, recognizing that enterprise AI success depends on strategy and implementation support, not just model capabilities.

A year after Meta tapped Alexandr Wang to build a new AI model, Zuckerberg has to sell it

infonews
industry
Jun 14, 2026

Meta spent $14.3 billion to hire Alexandr Wang and his team to build proprietary AI models, resulting in the Muse Spark model released in April 2024, a shift away from Meta's previous open-source approach. However, Meta still struggles to compete with OpenAI and Google, and faces challenges convincing developers and investors that it can monetize AI products beyond its core advertising business, which currently accounts for 98% of revenue. The company's earlier open-source Llama models failed to attract developers, damaging its reputation in the AI community.

Rethinking ransomware defense in the age of generative AI

inforesearchPeer-Reviewed
security

A Survey of Neural Network Robustness Assessment in Image Recognition

inforesearchPeer-Reviewed
research

Amazon security research reportedly led to the White House’s Anthropic Fable ban

infonews
securitypolicy
Previous182 / 485Next
research
Jun 15, 2026

Researchers found a security weakness in fingerprinting methods for image generation AI models (systems that embed hidden identifiers into generated images to protect ownership). Existing fingerprinting techniques are vulnerable to collusion attacks, where multiple attackers combine their models to remove or hide these fingerprints. The researchers propose a new fingerprinting method that encodes fingerprints into a personalized normalization module (PNM, a customization layer in the model) and uses anti-collusion mechanisms based on mathematical transformations that severely degrade image quality if models are combined, making colluded models unusable.

Fix: The proposed method introduces 'an anti-collusion mechanism based on lossless function-invariant parameter transformations' that 'significantly degrades the image generation quality of colluded models, making them effectively unusable.' Additionally, 'our method allows developers to efficiently create multiple copies of fingerprinted T2I models by reparameterizing the PNM without the need for retraining,' and the researchers 'introduce a worst-case optimization strategy to improve robustness against model-level attacks.'

IEEE Xplore (Security & AI Journals)
Jun 15, 2026

NewCore, an Israeli cybersecurity startup, has emerged from stealth mode with $66 million in funding to build an identity platform designed for the era of agentic AI (AI systems that can take autonomous actions). The platform uses Secure Split Key (SSK), a technique that prevents a specific class of attacks on SAML (Security Assertion Markup Language, a system for managing authentication) infrastructure, and includes features like hardware-bound credentials and continuous identity discovery to protect human, machine, and AI agent identities.

SecurityWeek
Jun 15, 2026

NewCore, a new cybersecurity startup, has raised $66 million to help companies manage AI agents as workplace participants by giving them digital identities with proper authentication and access controls. As companies increasingly deploy AI agents alongside human employees, NewCore argues that traditional identity platforms (systems that verify who users are and what they can access) are outdated and designed only for humans, not software workers. NewCore's platform treats AI agents as first-class identities with their own permissions and access controls, using a 'split-key' architecture (dividing credentials between the customer and the platform to prevent a single point of failure) to secure them.

Fix: NewCore's platform addresses this through several built-in features explicitly described in the source: a 'split-key' architecture that divides critical identity credentials between the customer and the platform to eliminate a single point of compromise; an 'Agentic Skill' integration package for coding assistants like Claude Code, OpenAI's Codex, and Cursor that allows AI tools to access enterprise systems as managed identities rather than through manually distributed credentials; and a mobile app that lets employees grant, review, and revoke access for AI agents, providing human oversight as companies deploy more autonomous systems.

TechCrunch (Security)

Fix: Microsoft addressed SearchLeak at the beginning of the month. With Microsoft having fixed CVE-2026-42824, there's no user action required to mitigate this threat.

BleepingComputer

Fix: Update Langflow to version 1.9.0 or later (current version is 1.10.0). The vulnerability affects versions up to 1.8.4, and the fix was released on April 15.

CSO Online
Jun 15, 2026

Anthropic stopped all access to its Fable 5 and Mythos 5 AI models after receiving an export control directive (a government order restricting who can use certain technology) that prevents foreign nationals from using these systems. The shutdown was triggered by US regulations that treat advanced AI as technology that needs restricted access.

Dark Reading
Jun 15, 2026

Researchers discovered that attackers can exploit AI agent guardrails (safety systems that check AI behavior) by inserting malicious content into documents, causing the security mechanisms to enter extended thinking loops that dramatically slow down or crash shared AI systems. This reasoning-extension DoS (denial-of-service, a type of attack that makes systems unavailable) attack targets the safety layer itself rather than trying to jailbreak the AI model, and it works across multiple AI frameworks and different LLM families. Unlike traditional attacks that try to produce unsafe outputs, this technique compromises availability by exhausting computational resources, with some systems experiencing slowdowns of up to 148 times normal speed.

CSO Online
Jun 15, 2026

Sovereign cloud (cloud infrastructure located in a specific country or region to comply with data residency laws) alone does not guarantee the control that enterprises expect over their AI workloads, despite regulatory pressure in Europe and increasing scrutiny in the US. The real control point lies in identity governance (managing who can access what resources and under what circumstances) and related infrastructure layers like encryption key management, access logging, and workload identity management, not just where data is physically stored.

CSO Online
Jun 15, 2026

AI agents (software systems that can read data, process external content, and take actions) now commonly have all three dangerous capabilities together, making them vulnerable to prompt injection (tricky instructions hidden in data that trick the AI into doing harmful things). Security experts like Meta recommend the 'Rule of Two,' which limits agents to only two of these three capabilities per session and requires human approval if all three are needed, but this framework has limitations and doesn't fully solve the problem.

Fix: Meta's security team published the 'Rule of Two' framework, which recommends agents satisfy no more than two of the three trifecta properties (access to private data, exposure to untrusted content, ability to communicate externally) in a single session, with human-in-the-loop approval required if all three are necessary. Simon Willison endorsed this framework as 'the best practical advice for building secure LLM-powered agent systems today.'

CSO Online
Jun 15, 2026

AI agents operate at machine speed across multiple systems, making traditional security models that grant access once at login insufficient for protecting modern infrastructure. CrowdStrike's Continuous Identity approach continuously evaluates identity, device, threat, and business context to grant, adjust, or revoke access in real time, with specific features for AI agents including verification based on SPIFFE standards (an identity framework), removal of standing privileges (permissions that remain active indefinitely), and immediate revocation when risk conditions change.

Fix: CrowdStrike provides Continuous Identity for AI Agents through Falcon Next-Gen Identity Security, which eliminates standing privileges and verifies trust for every agent action in real time using SPIFFE identity standards and the Shared Signals Framework. The system evaluates each action against the human user's and agent's entitlements and current security and business context, ensures agents cannot exceed the permissions of their human operator, preserves human identity and permissions when agents delegate to sub-agents, and immediately revokes access if context changes (such as new vulnerabilities or HR status changes). Additionally, CrowdStrike Falcon AI Detection and Response (AIDR) continuously inspects prompts and intent to detect permission misuse, triggering Continuous Identity to revoke access before damage occurs.

CrowdStrike Blog

Fix: Apply patch 91a31aac1b0f4dbc6b8bef9f6eff0b7912e0bc65. The vendor confirms the fix requires authentication (proof of identity) and the research_exports capability (a specific permission) to access research export endpoints.

NVD/CVE Database

Cisco Catalyst SD-WAN Manager has a path traversal vulnerability (a flaw where attackers can access files outside intended directories) that allows authenticated remote attackers to create or overwrite files on affected systems. This vulnerability is actively being exploited by attackers.

Fix: Apply mitigations according to Cisco vendor instructions and follow CISA's BOD 26-04 guidance for prioritizing security updates. If mitigations are unavailable for cloud services, discontinue use of the product. Stakeholders must evaluate each system's internet exposure and ensure compliance with BOD 26-04 patching guidelines by the due date of 2026-06-29.

CISA Known Exploited Vulnerabilities

The LiteSpeed cPanel plugin has a symlink following vulnerability (a flaw where the software unsafely follows symbolic links, which are shortcuts to files, allowing attackers to access unintended files) that affects shared hosting servers using CloudLinux/CageFS. An attacker with FTP or web shell access (the ability to run commands on a web server) could exploit this vulnerability, and it is currently being exploited in active attacks.

Fix: Apply mitigations in accordance with vendor instructions from the LiteSpeed security update (https://blog.litespeedtech.com/2026/06/01/security-update-for-litespeed-cpanel-plugin-2/), following CISA's BOD 26-04 guidance for patching. If mitigations are unavailable, discontinue use of the product. The patching deadline is 2026-06-18.

CISA Known Exploited Vulnerabilities
Jun 14, 2026

According to a report, the White House may have restricted exports of Anthropic's Mythos AI model because it feared a group linked to China had accessed it, which would pose serious national security risks. One concern is that the Chinese government could use distillation (training a simpler AI on a more advanced one to copy its behavior) to reverse engineer the model.

The Verge (AI)
OpenAI Blog
CNBC Technology
research
Jun 14, 2026

This article examines how ransomware (malicious software that locks files and demands payment to unlock them) defense strategies need to change as generative AI (AI systems that create new content like text or code) becomes more common. The piece suggests that traditional security approaches may be less effective in an environment where AI is widely used.

Elsevier Security Journals
safety
Jun 14, 2026

This academic survey paper reviews methods for testing how well neural networks (AI systems trained to recognize patterns in data) perform when faced with unexpected or manipulated images. The paper examines various approaches researchers use to assess whether image recognition systems remain accurate and reliable under challenging conditions.

ACM Digital Library (TOPS, DTRAP, CSUR)
Jun 13, 2026

Amazon's security research found that Anthropic's Fable 5 AI model could be manipulated through prompt injection (tricking an AI by hiding instructions in its input) to reveal information usable for cyberattacks. After Amazon CEO Andy Jassy shared these findings with the White House, Anthropic restricted access to Fable 5 and Mythos 5 to prevent foreign nationals from using the models.

The Verge (AI)