No Free Lunch Podcast - Are They Listening and AI Risk Assesment
AI assistants and connected devices can create real business risks when they collect, process, or expose information without adequate controls.
In This episode of the No Free Lunch Podcast, Greg Stewart chats to Richard Frost, Head of Technology Solutions and Consulting at Armata Cyber Security.The discussion’s central message is clear: AI assistants and connected devices can create real business risks when they collect, process, or expose information without adequate controls. Richard and Greg explore how organisations can use AI while protecting sensitive data, and why preparation matters as much as prevention.
Watch the Podcast Here:
Key discussion points
- AI is both a tool and an expanding attack surface. Richard describes how AI can help cybersecurity teams identify vulnerabilities and respond faster. But as businesses and employees adopt more AI tools, those systems also create new ways for information to be exposed or misused.
- Shadow IT and data leakage are major concerns. Employees may use AI services that their organizations have not approved or secured. Richard warns that prompts and integrations could expose sensitive business information—including client, financial, or CRM data—if access and usage are not carefully controlled.
- Everyday devices deserve attention. The conversation considers how phones, smartwatches, voice assistants, and smart TVs may collect information in the background. Greg and Richard discuss the difficulty of knowing what data is gathered, where it goes, and how it may be used.
- Policies should define acceptable AI use. Richard recommends setting clear rules: which AI tools employees can use, what information they can share, and how data should be stored. He also highlights data sovereignty—the importance of knowing where information is hosted and which privacy laws apply, including POPIA and GDPR.
- Personal devices make oversight harder. Workplace controls may not extend to an employee’s own phone or devices at home. That makes sensitive conversations, meeting recordings, and connected devices worth considering as part of an organization’s security planning.
- Resilience is essential—even when prevention fails. Richard argues that companies should prepare to limit the impact of a breach: contain it, protect other systems, and recover quickly. The discussion also raises the idea of a “kill switch” or other way to shut down a system if it behaves unexpectedly.
- AI raises broader governance questions. The episode closes by considering how quickly AI is developing, the risks of weak safeguards, and the potential for tools to be used in cyberattacks or to expose proprietary information. The takeaway for businesses is to approach AI thoughtfully, protect sensitive data, and plan for what to do when something goes wrong.
Was this useful?0 reactions
