20 May 18

Cultivating trust in Artificial Intelligence technologies

Trust is defined as ”a firm belief in the reliability, truth, or ability of someone or something,” and it is often attained through successful and repeated delivery of what has been agreed upon or is expected. Once lost, however, trust is an incredibly difficult thing to regain, and numerous corporations and individuals can attest to this position.

In today’s world, we’re seeing more people, and younger ones in particular, becoming increasing comfortable with placing higher levels of faith or trust in digital ecosystems and the transactions they enable in exchange for convenience. It’s a delicate relationship. Should that trust be abused or betrayed, the equation breaks down, in some cases irrevocably. This is why securing digitisation is one of the overriding imperatives of a sustainable future. As networks and end-user devices grow in complexity and number, so must the ability for systems to monitor and react to rising cyber threats effectively.

Security has to be pervasive without being invasive. Individuals will lose trust in digital services with loss of privacy, and automation will be held back by lack of safety. Facebook, for instance, is in the midst of one of the deepest crises of confidence in its history following the fallout from having granted a third-party data analytics company access to the personal details of tens of millions of user accounts. The billions of dollars already wiped off Facebook’s market valuation in the wake of the news may be overshadowed by the longer-term impact from lost trust in Facebook’s ability or conviction to maintain the integrity of users’ personal data.

The exponential growth in compute power, storage, and connectivity, matched by the unrelenting rise of threat vectors and actors in parallel, makes it unrealistic to expect legacy methods of cyber security to keep pace with the rise of the Internet of Things (IoT). Evolving threats such as the infiltration of the electric grid by hackers, or the hacking of vehicle electronics are real and rising.

As we enter an age of pervasive/ambient computing, Machine Learning forms just one component of Artificial Intelligence, which could have a drastic and lasting impact on the ability for the system to establish cyber security resilience. However, for this to become sustainable, the application of security should be end-to-end, or else the weakest link will give way.

In a positive development for digital security, it is becoming widely accepted that for IoT to work optimally and become sustainable, it requires proactive protection, with end-to-end resilience being paramount. Resilience here refers to the ability of digital systems to face compromise, recover and restore functionality with minimal disruption.

Innovation is unrelenting. The advent of 5G networks will drive the rise of truly smart cities and nations, with the further enablement of cloud-based interactions supporting the explosion of sensors and digital connections. These interconnected environments are brimming with potential, and to reap the full benefits of digital transformation the guiding principle must be the institution of digital security and perpetuation of trust in a step-by-step process.

What are the necessary tools to establish this dome of trust? We believe the implementation of stronger cryptographic algorithms (post-quantum crypto) and the use of Public Key Infrastructure is central. As is the attainment of cyber resilience through the implementation of end-to-end digital security and applications such as Blockchain and Secure Communications at all levels including hardware, software, applications, and voice/data channels.

Securing digitisation is not a one-tool solution or a one-time fix. It has to be end-to-end, not merely in one component. It requires ongoing consideration and implementation, taking place incrementally as technology advances, and with the end-goal being that digital security becomes indistinguishable and indivisible from the ecosystems it is implemented to protect.

About the author

Ticky Thakkar is Chief Scientist – Advanced Research at DarkMatter, where he is responsible for the development of key technologies and research for the firm’s new generation products. He may be contacted on Twitter @TickyThakkar

By Ticky Thakkar
  Back To Blog Listing