16 May 2025
Can We Trust AI? Cybersecurity and Manufacturing with FISC
Artificial intelligence promises efficiency, adaptability, and smarter production. But can AI systems really be trusted in manufacturing environments where safety, reliability, and transparency are non-negotiable?
In this episode of the MPD Q, A & I podcast, Peter Sund, CEO of FISC (Finnish Information Security Cluster), a branch of the Technology Industries of Finland, raises the critical question of whether AI systems can be trusted in manufacturing. His answer is nuanced: no and yes, if the right control measures are applied.
The Three Dimensions of Trust in AI
Sund breaks down the issue into three interconnected layers:
1. Cybersecurity of AI
AI systems are computer systems that process data, and like any digital infrastructure, they must be protected to ensure confidentiality, availability, and data integrity. This includes securing the data, models, and platforms involved.
2. Security by AI
AI can also be an asset in cybersecurity. In Sund’s words, much of cybersecurity is about processing large volumes of data, something AI excels at. In this role, AI helps detect anomalies, automate monitoring, and respond to digital threats efficiently.
3. Cyber Insecurity from AI
AI can also be weaponized. Malicious actors use AI to conduct sabotage, espionage, and financial crimes. A striking example is the discovery of thousands of leaked API keys embedded in the Chinese DeepSeek model, highlighting how model training data can expose sensitive digital assets.
A Stark Warning: Security Is Your Responsibility
“Unfortunately, the security authorities will not provide cybersecurity for manufacturing companies, they have to do it by themselves,” says Sund.
Because digital systems and data are private property, the responsibility lies with the owner.
“Like your bike, car or summer cottage, you are responsible for taking care of it,” he adds.
Moreover, Sund observes that despite decades of digital innovation, there’s been little meaningful progress in state-level cybersecurity capabilities. Organizations must act independently to stay secure.
AI Helps the Good Guys, and the Bad
AI increases productivity not only for industry leaders, but also for criminals. Sund warns that AI is enabling bad actors to scale cyberattacks, create new forms of sabotage, and even disrupt industrial operations.
He also notes the irony that AI, initially intended to automate mundane tasks, is now being used to generate music, poetry, and art, things we once considered deeply human.
That duality is the challenge: a powerful tool that can be used for both progress and harm.
The Business Case for Cybersecurity
While cybersecurity is often seen as a cost, Sund reframes it as a business-critical investment, essential for maintaining customer trust and ensuring operational continuity.
Cybersecurity is often seen as a cost.
“Committing investments in a systematic and consistent way, with a view of projected costs, is much better than facing 10, 50, or even 100 times the cost later, plus other negative consequences, suddenly and without any plans,” Sund says.
His final point: risk should never be accepted by default. Include cybersecurity in your broader business risk assessments, so that even if you accept risk, it’s a conscious and informed decision.
The MPD Q, A & I podcast series, part of the Manufacturing Performance Days, dives deep into the role of AI in creating new digital business models, advancing technologies, and leading sustainable transformation. Through candid conversations with leading professionals at the forefront of AI transformation, the series offers listeners first-hand insights into the future of AI-powered manufacturing.
Find the full episode on
Spotify
https://open.spotify.com/episode/2fBp5iPo9sqKJGf9F3T2Do?si=7077fe6a78684ae8
Apple Podcasts