SaaS Security | Breaking Cybersecurity News | The Hacker News

AI-powered browsers are changing how we use the web, but they're also creating some serious new security risks. Tools like Perplexity's Comet and Opera's Neon can summarize pages and automate tasks for you. The problem is that researchers have found these agentic copilots can be hijacked by malicious prompts hidden in ordinary webpages, essentially turning your browser against you. In August 2025, Brave's security team disclosed an indirect prompt injection against Perplexity's Comet using hidden instructions in a Reddit spoiler tag, leading Comet to extract an email address and a one-time passcode. No memory corruption, no code execution exploit. The browser simply followed instructions it couldn't distinguish from legitimate user intent. In this post, we'll look at how these attacks work, why they slip past traditional defenses, and what security teams can do to keep data safe from compromised AI agents. AI Browsers: Powerful, But a New Target AI-ena...

Enterprise adoption of AI is no longer a future trend; it's a present-day reality. As organizations race to leverage AI for innovations, security teams are grappling with a new, complex, and dynamic attack surface. AI is breaking the operational silos that currently segregate Cloud, SaaS and Endpoint Security; AI is everywhere and it is consuming enterprise data and assets across these channels. Traditional security tools, designed for cloud infrastructure and SaaS applications, are fundamentally ill-equipped to handle the unique risks posed by AI.  AI security posture management (AI-SPM) solutions can provide relief by protecting critical AI assets, but it's important to note that not all AI-SPM solutions are created equal. Many solutions offer only basic posture checks and are focused predominantly on infrastructure and vulnerability management. In addition, most focus solely on Cloud or SaaS, leaving many blind spots when trying to get the full picture of your AI landscape. ...

In November 2025, Anthropic revealed a cyber espionage campaign dubbed GTG-1002, the first documented case of an AI agent orchestrating real-world intrusions with minimal human input. A Chinese state-sponsored group manipulated Anthropic's Claude Code assistant into executing about 80% of a multi-target hacking campaign autonomously. Instead of merely advising cybercriminals, the AI took control of key steps: reconnaissance, vulnerability discovery, exploitation, credential theft, and data exfiltration across dozens of organizations. The result was an operation running at machine tempo. Claude performed tasks in a fraction of the time a human team would need, even identifying sensitive databases and writing exploits in seconds. Figure 1: The distinct phases of the Claude cyberattack At the peak of the attack, the AI made thousands of requests (often several per second), an onslaught of activity impossible for humans to match. This speed and scale of automation is a game changer: a...

In cybersecurity, the old adage "trust but verify" emphasizes that granting trust should always be accompanied by oversight. Yet, with software-as-a-service (SaaS), organizations often stop at the "trust" part and never get around to the "verify." SaaS environments in 2025 run on implicit trust. Once a user or app is authenticated and given access, it's largely trusted indefinitely. Tokens issued to third-party apps rarely expire, integrations often get more permissions than they truly need, and automations execute with minimal human oversight. We talk about Zero Trust principles, but in practice, many SaaS platforms grant one-time approval and then assume all is well thereafter. The result is a growing security gap, where credentials and connections are implicitly trusted far beyond what's safe, creating fertile ground for breaches and abuse. Implicit Trust in the SaaS Ecosystem Every SaaS integration or API token represents an implicit trust relationship between your organizatio...

As SaaS ecosystems expand, not every user is human anymore. AI assistants, automation bots, integration services, and API tokens now perform countless actions across business cloud applications, often with the same or greater access privileges as employees. These non-human identities (NHIs) are silently driving productivity while introducing a new class of risk: unmonitored, long-lived, and often misunderstood access. These machine credentials (service accounts, API keys, OAuth tokens, etc.) are essential for automation and integrations, but their growth far outpaces the oversight and security controls applied to them. The result is a widening visibility gap. A lot of NHI types enjoy broad permissions within SaaS apps, sometimes more privileges than a human user, yet they rarely get the same scrutiny as employee accounts. Over-privilege is common: about one-third of SaaS app integrations have access to sensitive data that exceeds their needs. Let's examine a few notable data brea...

ShinyHunters is a notorious cybercrime group that has resurfaced with a new playbook of SaaS-focused attacks. Known for monetizing stolen data on underground forums since 2020, ShinyHunters has historically breached companies by stealing credentials and databases. Recently, however, they've shifted tactics to aggressive social engineering, mirroring the methodology of the Scattered Spider group. Instead of exploiting software vulnerabilities, ShinyHunters now exploits human trust, targeting the underbelly of third-party SaaS platforms through impersonation and phishing. In mid-2025, a wave of breaches struck companies like Google, Workday, Pandora, Cisco, Chanel, and others, all tied together by one common thread: the attackers leveraged access to these firms' Salesforce CRM or similar cloud systems. Below, we look at what happened in the Google and Workday breaches, examine techniques ShinyHunters used, and demonstrate how a dynamic SaaS security approach (like Reco's) could have...

Device and software vulnerabilities pose an increasing risk to modern security. However, patch management is an infamously difficult (and downright Sisyphean) task for IT and security teams, who are faced with an ever-growing list of CVEs to remediate. This task was difficult enough in the days of on-premise environments, but a modern distributed workforce has to contend with all the users, devices, and applications that may exist outside the purview of traditional security solutions, like MDM. Overall, with the ever-growing number of CVEs and the ever-growing sprawl of shadow IT, patch management has become both more urgent and more daunting than ever. IT and security teams need to adopt zero trust methods to ensure that only healthy and patched devices are able to access their critical systems. With the help of SaaS management and employee-remediation tactics, teams can do even more to improve efficacy and support for their company-wide patch management programs.  French philo...

Generative AI has quietly become a part of the SaaS ecosystem that businesses use every day. Platforms like Zoom, Slack, Microsoft 365, and Salesforce now have AI assistants. You can use these tools to do things like write summaries of meetings or perform routine tasks. A recent  survey found that 95% of U.S. businesses now use generative AI. This is a big increase from last year. But this quick growth of AI features is making security leaders worried. Sensitive information could be leaked or used in the wrong way if there aren't enough controls in place. Shadow AI and Its Far-Reaching Risks When employees use AI apps without the knowledge or approval of IT, it creates shadow AI . This is akin to the shadow IT problem of unsanctioned cloud apps, but now with AI services. The unauthorized use of AI platforms can unknowingly expose organizations to data privacy issues, compliance violations, and even disinformation risks. We're already seeing these risks play out. Samsung engin...

When generative AI tools became widely available in late 2022, it wasn't just technologists who paid attention. Employees across all industries immediately recognized the potential of generative AI to boost productivity, streamline communication and accelerate work. Like so many waves of consumer-first IT innovation before it—file sharing, cloud storage and collaboration platforms—AI landed in the enterprise not through official channels, but through the hands of employees eager to work smarter. Faced with the risk of sensitive data being fed into public AI interfaces, many organizations responded with urgency and force: They blocked access. While understandable as an initial defensive measure, blocking public AI apps is not a long-term strategy—it's a stopgap. And in most cases, it's not even effective. Shadow AI: The Unseen Risk The Zscaler ThreatLabz team has been tracking AI and machine learning (ML) traffic across enterprises, and the numbers tell a compelling story. In 2024 ...