A naukri.com initiative
Cybersecurity-Insiders
1M
314
Image Credit: Cybersecurity-Insiders
Creating a Copilot That Doesn’t Violate Security and Compliance
- Low code/no code development is enabling business users to build apps and automations throughout the enterprise, with tools like Microsoft Copilot Studio allowing users to build copilots and AI agents.
- However, with this new technology comes new risks, and security teams must establish guardrails to prevent data leaks and security backdoors and ensure AI does not act out ofbounds.
- Common risks with low code/no code include overprovisioning access and embedded credentials, and there is a lack of security visibility as to who is building what and the ensuing risks.
- Organisations must consider how to spot “bad activity” if AI gains access to corporate data and make sure it isn’t accessing what it shouldn’t internally.
- IT and security teams need better visibility into what people are building with AI apps since this is now happening outside of traditional IT parameters.
- Monitoring and scanning tools are essential and teams should erect stringent guardrails on the back end so sensitive data doesn’t get overshared while fostering innovation.
- AI agents and copilots acting autonomously are a huge lift for security to protect the enterprise from data leaks and cyber-attacks.
- IT and security teams need visibility, monitoring, and controls for businesses to flourish while keeping their data safe.
Read Full Article
18 Likes
For uninterrupted reading, download the app