Please consider supporting us by disabling your content blocker.
loader

Tech Summit Image

AI & Its Limitations in Cybersecurity

At the Tech.eu Summit in London, Dr. Ken Urquhart, Global Vice-President of 5G/Edge/Satellite at Zscaler, and Steve Brierley, Founder and CEO of Riverlane, discussed the critical intersection of artificial intelligence (AI), cybersecurity and quantum computing. Moderated by Duygu Oktem Clark, Managing Partner at DO Venture Partners, the talk underlined both the challenges and opportunities these technologies present.

Urquhart opened the discussion by addressing the limitations of AI in cybersecurity.

“AI, as we apply it today, involves algorithms that are interpretable and useful for cyber defense,” he said. However, he pointed out that current AI technologies, such as neural networks and large language models, come with issues like statistical variability and hallucinations, where the AI “makes things up that may not be true.”

Urquhart explained that these statistical models could become less accurate over time, adding: “You need to be thoughtful about how you apply AI because it can give less accurate answers if asked the same question twice in a row over a span of hours or days.”

Quantum Computing’s Potential & Challenges

Brierley shared his thoughts into the advancements in quantum computing and its implications for cybersecurity. He noted that while today’s quantum computers are “extremely error-prone” and capable of only about 100 to 1,000 operations before failure, significant progress is being made with quantum error correction.

“Quantum error correction is a layer that sits on top of the physical qubits and corrects errors in real-time,” Brierley explained.

This development is crucial for achieving cryptographically relevant quantum computing capabilities.

“2023 and 2024 have been pivotal years as we crossed the threshold in various qubit modalities, making error correction viable,” he said. Brierley projected that within the next two to three years, we could see quantum computers performing up to a million operations, surpassing what classical computers can simulate.

Ethical & Security Considerations

As AI and quantum computing advance, ethical and security challenges emerge. Urquhart stressed the importance of understanding AI’s current limitations.

“We are on a journey with artificial intelligence. It does not think; it is a collection of statistical outcomes,” he stated. Urquhart warned against over-reliance on AI for critical decisions, as its current form can lead to significant errors.

Brierley added that quantum computing has the potential to revolutionize industries, particularly in simulating molecular dynamics and chemical interactions.

“Quantum computers can replace time-consuming lab experiments with simulations, transforming industries like drug discovery and material science,” he said.

Collaboration for a Secure Digital Future

Both experts agreed on the necessity of collaboration among academia, industry and government to harness these technologies responsibly. Brierley called attention to the importance of a coordinated effort, likening it to a “Manhattan-scale project” to build the world’s most powerful quantum computers. “We need effective collaboration across sectors to ensure the technology benefits society,” he said.

Urquhart echoed this sentiment, giving emphasis to the role of commercial entities in driving innovation and the government’s role in providing a regulatory and funding environment.

“The machinery is there; we just need the will to engage and make it run,” he remarked.

The Future of Quantum & Cybersecurity

Looking ahead, both Urquhart and Brierley stressed the urgency of preparing for the impact of quantum computing on cybersecurity.

“Quantum computing will break most encryption at some point,” Urquhart warned, urging businesses to act now to mitigate future risks.

Brierley concluded: “Quantum computers are not just faster computers; they represent a massive step forward for specific problems, and their potential for both good and bad is immense.”

The discussion underscored the transformative potential of AI and quantum computing while cautioning against complacency. As these technologies evolve, proactive collaboration and ethical considerations will be paramount in shaping a secure digital future.

Featured image: Credit: Tech.eu