Aggregator
CVE-2022-32082 | MariaDB 10.5/10.6/10.7.0 dict0dict.cc get_ref_count assertion (MDEV-26433 / Nessus ID 252498)
CVE-2020-14547 | Oracle MySQL Server up to 5.7.30/8.0.20 Optimizer denial of service (Nessus ID 252499)
CVE-2023-28882 | Trustwave ModSecurity 3.0.5/3.0.6/3.0.7/3.0.8 Configuration Transaction denial of service (Nessus ID 252503)
Voyager 2 промахнулся — Уэбб разглядел скрытую луну Урана спустя 40 лет
Legitimate Chrome VPN with 100K+ Installs Secretly Captures Screenshots and Exfiltrates Sensitive Data
A Chrome extension marketed as FreeVPN.One, boasting over 100,000 installations, a verified badge, and featured placement in the Chrome Web Store, has been exposed as spyware that silently captures screenshots of users’ browsing activities and exfiltrates them to remote servers. Despite its privacy policy explicitly stating that the developer does not collect or use user […]
The post Legitimate Chrome VPN with 100K+ Installs Secretly Captures Screenshots and Exfiltrates Sensitive Data appeared first on GBHackers Security | #1 Globally Trusted Cyber Security News Platform.
The New Frontier: Why You Can’t Secure AI Without Securing APIs
The release of a new KuppingerCole Leadership Compass is always a significant event for the cybersecurity industry, offering a vendor-neutral view of the market's current state. The 2025 edition, focusing on API Security and Management, is critical as it arrives at a pivotal moment for technology. It clearly presents a fact many organizations are just beginning to understand: the crucial connection between the rise of Artificial Intelligence and the necessity for robust API security.
This is the first part of a three-installment blog series highlighting the main findings of this landmark report. In this post, we will emphasize its core theme: the interconnected and vulnerable relationship between AI and APIs.
In the new KuppingerCole report, analyst Alexei Balaganski explains that APIs have evolved far beyond simple technical tools; they now orchestrate business logic and drive automation across the entire enterprise. The rise of artificial intelligence has supercharged this trend. The analysts present a key finding that establishes the modern relationship between these technologies: "APIs are the backbone of Al: Every LLM integration, agentic Al workflow, or autonomous decision system depends on API calls". This fundamental shift means that APIs no longer just support business operations—they now actively define them.
This deep, API-driven integration introduces a significant and high-stakes attack surface. The report issues a stark warning about this new reality, explaining that any effort to secure an AI model itself is ultimately ineffective if its underlying connections are vulnerable. As the analyst puts it, protecting a model "is futile if the APIs that interface with those models are left unguarded". This vulnerability exposes organizations to a new class of AI-related threats, including prompt injection and data exfiltration, which are often executed through sophisticated business logic attacks that exploit an API's intended functionality to bypass traditional defenses.
This challenge is precisely what Salt Security was created to address. Our platform aims to look beyond common vulnerabilities and understand the specific logic and context of each API. KuppingerCole highlights our “patented AI/ML engine”, which it says “differentiates between benign anomalies and actual attacks with a claimed 92% intent accuracy”. This capability is essential for identifying sophisticated, low and slow attacks targeting business logic, which AI-driven threats often exploit.
The report also supports our strategic approach, noting Salt's early efforts in AI security by providing protections against prompt injection and other threats specific to LLMs. As you develop your AI strategy, securing the APIs that connect these advanced models to your vital data is not just recommended; it’s essential.
With an understanding of the AI-driven threat landscape, our next post will explore what it takes to lead in this challenging area and why KuppingerCole recognized Salt Security as a clear Overall Leader.
The insights from the KuppingerCole report provide a clear roadmap for navigating this new, AI-driven threat landscape. To see the full, independent analysis and understand why Salt Security was named an Overall Leader, download your complimentary copy of the report today. And when you’re ready to move from strategy to action, we invite you to take the next step with our free, personalized API Attack Surface Assessment to discover and prioritize the specific risks within your own environment.
The post The New Frontier: Why You Can’t Secure AI Without Securing APIs appeared first on Security Boulevard.