The ripple effects of Artificial Intelligence

Is Artificial Intelligence (AI) an advantage to your business? While most say yes, tech experts concur that AI brings in a whole new set of challenges to organizations, especially around data security, transparency and data protection.

Irrespective of the challenges, analysts have predicted that the artificial intelligence market will register a CAGR of over 33% by 2023. But, it’s time that organizations start addressing a few key aspects around data.

Ajoy Menon, Managing Director and Lead – Financial Services, Accenture Advanced Technology Centers in India, informs, “We are seeing great traction among BFSI companies to incorporate AI along the whole spectrum of its value chain and to reimagine their processes, products, employees and customer experience.” He, however, warns that financial institutions must focus on building a deeper understanding on when or how these AI systems can go wrong.

“As AI systems are typically integrated into a broader ecosystem of technology and human processes, robustness and reliability of these systems, data manipulation and theft, and lack of transparency have become the top security and safety issues facing AI systems today. Organizations need to “raise” responsible AI systems and establish a comprehensive Data protection and privacy framework to create trust, thereby ensuring the security and dependability of AI systems.”

Interestingly, AI is touted as a technology that will, in many ways, act as a great tool for cybersecurity experts in the coming days. But there are two sides to this story, experts argue.

“Implementation of AI is slowly reaching its zenith, and in one way or the other, AI and security were made to help in the success of the other. However, organisations when dealing with AI-based applications deal with many external intrusions to obtain data. Apart from this, hacking is becoming more sophisticated as hackers have taken to AI to help in phishing and cybercrimes. The only clear-cut way to counteract against these threats is to create machine-learning algorithms, which detect abnormal activities carried out by unethical hackers,” says Anand Subramaniam, Head, Artificial Intelligence Practice at Aspire Systems.

To address these concerns, Subramaniam believes, that data Protection should be taken care of across three levels – data authorization and authentication, segregation of duties and application-based access.

“Apart from this, inbound and outbound data should always be encrypted. There should also be a control of peripherals, such as copy, pasting and printing options,” he adds.

Once each of these grounds is covered organizations should aim at data integrity being the priority. With incorrect, inconsistent and incomplete data, AI and ML act provide actionable intelligence. “Even though data integrity does not fall under data protection, it rather represents the same end goal. Ensuring such a tight-knit security process will ensure data is protected across each level to prevent cybercrimes and hacking attempts,” Subramaniam sums up.