Manish Mimani, Founder and Chief Executive Officer of Protectt.ai, discusses the strategic investment in AIShield, plans to build a full-stack AI security platform and the growing need for AI governance across regulated industries
Protectt.ai is strengthening its position in the rapidly evolving AI cybersecurity market through a majority investment in AIShield, an AI security platform incubated within Bosch Global Software Technologies (BGSW). The strategic partnership brings together Protectt.ai’s expertise in mobile application security and established relationships with regulated enterprises alongside AIShield’s research-driven capabilities in AI model assessment, red teaming, runtime protection and safety guardrails.
The investment marks an important step towards building a comprehensive AI security offering as enterprises move beyond experimentation and integrate artificial intelligence into critical business operations. With more than 40 patent filings and 11 granted patent families, AIShield brings intellectual property and engineering expertise to the partnership, while Protectt.ai provides a foundation for expanding its commercial reach across banking, financial services, manufacturing, automotive and government sectors.
The combined strategy places particular emphasis on securing AI systems throughout their lifecycle, addressing vulnerabilities in models, applications and autonomous agents while strengthening enterprise visibility, governance and operational control. The Middle East, where organisations are accelerating AI adoption amid growing security and regulatory expectations, represents an important market for the company’s expansion plans.
Manish Mimani, Founder and Chief Executive Officer of Protectt.ai, discusses the rationale behind the investment, the practical implications of full-stack AI security, the metrics enterprises should use to assess protection effectiveness and the company’s ambitions for the UAE and wider Middle East market.
Interview excerpts
What drove Protectt.ai’s majority investment in AIShield, and which product and market priorities will the new capital support?
We were already building AI security capabilities because we could see how quickly banks and other enterprises were adopting AI. Customer-facing chatbots, voicebots and internal agents create security requirements that go beyond protecting a mobile application. We wanted to build a business focused on that requirement.
AIShield brings research and technology developed within Bosch, while Protectt.ai brings experience working with global banks and other regulated organisations. Bringing those capabilities together gives us a stronger basis for building a full-stack AI security business. That covers both securing AI systems and using AI to identify vulnerabilities in applications and infrastructure.
The capital will support product development, research and development, and global market expansion, including the development of globally trusted AI security platform.
Banking and financial services are a key starting point because that is where we have strong customer relationships. Automotive, manufacturing and government are also important markets.
The priority is to translate the combined research and product capabilities into solutions that enterprises can use as AI becomes part of their day-to-day operations.
How will combining mobile application security with AIShield’s capabilities help enterprises secure AI from development through production?
A bank can have a secure mobile application and still face risks in the AI services behind it. For example, a customer may interact with a chatbot, while AI agents support onboarding or internal processes. Those models and agents need their own security controls. AIShield addresses that enterprise AI environment; it is not limited to mobile devices.
Before deployment, model scanning helps identify vulnerabilities and possible backdoors. Red teaming tests how an AI application responds to attempts to manipulate its behaviour. In production, runtime protection and guardrails check inputs and outputs, helping to filter malicious prompts and inappropriate responses.
These controls need to continue after a model is approved. Models should be tested regularly, with runtime policies used to mask or block sensitive information where required.
Our aim is to bring that AI security capability together with our mobile application security experience. The two address different parts of an enterprise’s technology environment.
We want customers to be able to secure the application they offer users and the AI systems supporting it, from assessment through day-to-day use.
Which AI security gaps are most pressing for banks and regulated enterprises, and how will you measure the combined platform’s effectiveness?
For banks and regulated enterprises, the most pressing AI security gaps include limited visibility into AI model supply chains, vulnerabilities in model files and dependencies, adversarial attacks, sensitive data exposure, and inadequate runtime protection. Addressing these risks requires continuous assessment, automated red teaming, and real-time security enforcement.
The combined platform’s effectiveness will be measured through key security metrics. AI Bill of Materials (AI BoM) will provide visibility into model components and supply-chain risks. Vulnerability Discovery Rate and Attack Scenario Coverage will measure security findings and the breadth of threats tested. Repeat red-team assessments will help evaluate improvements in AI security posture.
For runtime security, Threat Detection Rate, Attack Prevention Rate, and Policy Enforcement Success will assess threat identification, attack prevention, and enforcement accuracy. False Positive Rate will measure disruptions to legitimate requests, while Sensitive Data Protection will evaluate the prevention of exposure involving PII, credentials, and financial information.
Together, these metrics demonstrate measurable improvements in AI security and risk reduction.
What does Full-Stack AI security mean in practical terms, and how will your offering address UAE and Middle East enterprises’ control and governance requirements?
Full-Stack AI Security means protecting the entire AI ecosystem—from models and applications to AI agents and their interactions with enterprise systems. As organisations accelerate AI adoption, securing these interconnected layers is essential to prevent vulnerabilities, data exposure and emerging cyber threats. As organisations across global markets, particularly the Middle East and Africa (MEA), accelerate AI adoption, securing these interconnected layers is critical to mitigating vulnerabilities, preventing data exposure and defending against emerging cyber threats.
Our approach combines AI Model Scanning, AI Red Teaming, Runtime Guardrails and Agentic AI Security to identify weaknesses, test resilience and strengthen protection across the AI lifecycle. We focus on helping enterprises adopt and scale AI with greater confidence, supported by security capabilities and intellectual property developed in-house.
We are working towards bringing AIShield’s offerings to the UAE market. The focus is on helping enterprises secure the AI they adopt and strengthen governance as they scale its use. Specific deployment and control requirements would need to be established with each organisation; they should not be assumed to be identical across the region.
Source: Tahawul Tech