what new compliance challenges?

AI, digital identity control and process-mining are increasingly inseparable concepts as regulations around AI begin to multiply and companies will have to improve their compliance in this area…

According to a recent study by Capgemini, 78% of individuals want to use AI-based technology solutions, especially for managing their digital identity and their interactions with traditional financial services and “Big Tech” payment platforms.

The pandemic has accelerated the adoption of chatbots and other virtual assistants, including when registering for services. This trend has thus amplified the needs in terms of biometric identity verification which, beyond streamlining the processes during registration, reinforce the level of security.

This is confirmed by a survey by Payments.com which reveals the flaws associated with the use of passwords: nearly 3/4 (75%) of users “rely on their memory to remember them” and 90% use the same on several sites.

Major challenges to overcome

Despite the inherent benefits of AI recognized by individuals, a number of challenges remain. The main one being the persistent mistrust of some users regarding the impact of these technologies on respect for their privacy. Research from Paysafe showed that a third of respondents would share their biometric information more easily if providers were more transparent about how they collect, manage and secure it.

Another major challenge is related to the need for users to have a better control over their digital identities. This management is usually a centralized process and users fear being prone to cyberattacks and other privacy breaches. A trend is currently emerging towards a more decentralized management framework, giving users greater control, without having to rely on third-party service providers.

Supervision thanks to European and international regulations

According to the OECD report on national AI policies, more than 700 regulatory initiatives in this area are being developed in more than 60 countries, but none of them has so far obligatory. There are, however, charters and ethical principles for AI, enacted by international standardization bodies, and intended to minimize the risks of these technologies so that they arouse the confidence of users.

AI regulation is inevitableas evidenced by the draft law on artificial intelligence proposed by the European Commission (Editor’s note: The European “AI Act”). This proposes to impose risk-related obligations on AI providers to protect users’ rights, while promoting innovation and the economic opportunities associated with these technologies. Government institutions are therefore strengthening their regulatory capacities in order to deal with the proliferation of AI applications which may, albeit unintentionally, lead to discriminatory situations. More than just a wish, AI governance is now a requirement and an imperative. The regulations will therefore impose more stringent compliance obligations on companies, backed by binding enforcement mechanisms such as penalties for non-compliance.

But adherence to trusted AI should not be viewed solely through the lens of risk mitigation. It is also about business opportunities because trust promotes customer satisfaction and loyalty. It is therefore a solid competitive advantage.

Ensure the compliance of its AI through reliable means

Businesses can ensure compliance of AI-enabled technologies while avoiding reducing business value in three ways. First, by taking a data-driven approach, to determine where applications of AI technologies can have the most impact, before deploying them. Is it to improve customer engagement, gain operational efficiency or mitigate compliance risk? Each of these factors requires an understanding of how these processes perform.

Also according to the Paysafe study, 56% of companies consider collecting data from borrowers to be one of the most difficult and inefficient steps in the loan application process, resulting in high dropout. Although proven to improve the efficiency of the loan application process, biometric identification and AI-based data collection technologies are also likely to create compliance risks, particularly in this area. regarding data privacy and algorithmic bias. To mitigate and address these risks, Task Mining enables organizations to collect information associated with underwriting processes to determine whether they comply with recommendations regarding the use of biometric identification and user-driven data collection systems. the AI.

On the other hand, the combination of Task Mining a you Process Mining can also help companies better understand how processes are carried out by visualizing its various stages and identifying delays, blockages and parasitic values. Process Mining provides factual and quantitative data from which companies can support their decisions, evaluate automation opportunities, monitor the performance of AI systems and feed Machine Learning. They will also be able to have predictive information about the future state of the processes in order to anticipate measures aimed at optimizing the results.

Furthermore, once equipped with data-driven insights to identify automation opportunities, companies can further accelerate the implementation of more advanced IT solutions.Intelligent Process Automation. AI-enabled applications are able to automate document-related processes, especially those that are labor-intensive and error-prone (mandates and compliance audits, KYC or AML in financial services but also supply chain logistics and Purchase to Pay and Order to Cash processes). These capabilities allow organizations to automate up to 95% of the document management process, cut development time in half, and reduce repetitive tasks.

Compliance helps organizations proactively navigate a complex and rapidly changing regulatory framework. It also fosters trust in AI both internally and with their customers and users. This is why proper AI governance must be embedded at all levels of a business, with clear channels to communicate and respond to potential risks.

By Andre Perryethics evangelistABBYY

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *