Mobile security firm Appdome has launched new dynamic defense plugins to detect and counter threats from DeepSeek AI, a controversial chatbot app accused of unauthorised data collection and surveillance risks. The new security feature, available on Appdome’s AI Native Defense platform, is designed to safeguard mobile enterprise apps, remote access, and corporate data from potential DeepSeek spyware attacks.
DeepSeek, a free AI-powered chatbot, has surged in popularity but raised alarms over its data security practices. Reports suggest the app can harvest user data and transmit it to external servers, particularly in China, without the user’s knowledge. Concerns over its risks have prompted government agencies in the United States and South Korea to propose bans on DeepSeek in official work environments. However, enforcement remains challenging, particularly in Bring Your Own Device (BYOD) settings where employers lack direct control over installed apps.
“The explosive popularity of DeepSeek AI creates a serious and immediate risk to mobile enterprises,” said Tom Tovar, Co-Creator and CEO of Appdome. “If you’re concerned about DeepSeek infiltrating your apps or exposing your data, we now offer a way to detect and prevent it.”
Proactive Defense Against DeepSeek Threats
Appdome’s new plugins leverage behavioural analytics to detect suspicious activities linked to DeepSeek, such as unusual file access, data extraction, user monitoring, and network traffic to external AI servers. The solution provides enterprises with multiple enforcement options, allowing them to either block DeepSeek’s operations, alert administrators, or take other defensive measures to protect their mobile ecosystems.
Unlike traditional mobile security solutions, Appdome’s no-code approach allows businesses to integrate DeepSeek detection without SDKs, manual coding, or server deployment. This means enterprises can rapidly implement protections across their mobile apps without disrupting operations.
Chris Roeckl, Chief Product Officer at Appdome, highlighted the growing challenge posed by AI-based tools: “Today, the threat comes from DeepSeek. Tomorrow, it could be another AI tool. Cybersecurity teams need the ability to detect and enforce policies on AI use responsibly.”
DeepSeek’s Security Risks Under Scrutiny
Independent security analyses have revealed multiple vulnerabilities in DeepSeek’s architecture. The chatbot app has been found to:
- Transmit user data without proper encryption, exposing users to unauthorised data collection.
- Use outdated cryptographic algorithms, making it vulnerable to reverse engineering.
- Lack anti-tampering protections, allowing attackers to inject malicious code and bypass security controls.
- Enable runtime memory extraction, which could expose sensitive data such as authentication tokens and cryptographic keys.
Researchers also found “jailbreaks” that circumvent DeepSeek’s built-in security restrictions, making it possible for attackers to manipulate its AI model for unintended purposes.
Kai Kenan, VP of Cyber Research at Appdome, stressed the need for vigilance: “Even informed users can unknowingly download DeepSeek AI onto BYOD or shared devices used in highly sensitive operations. With Appdome’s DeepSeek detection, organisations no longer have to be in the dark about the risks.”
A New Layer of Security for the Enterprise
The introduction of Appdome’s Detect DeepSeek Attack plugins marks a crucial step in addressing AI-related security threats. The solution is particularly relevant for businesses with enterprise apps, remote work environments, and mobile BYOD strategies, where traditional app bans may be difficult to enforce.
With DeepSeek and similar AI tools gaining traction, enterprises face increasing pressure to monitor, detect, and mitigate AI-driven risks before they compromise sensitive data.
