TrustVector
Building AI Trustworthiness through AI Governance and Solution Assessments

“Inability to easily determine trustworthiness of an artificial intelligence (AI) solution is a monumental problem impacting all levels of our society, ” says Aleksandar Jevtic, co-founder and CEO of TrustVector.

To put things into perspective, let’s consider an artificial intelligence (AI) tool built to detect melanoma cancer by analyzing a new, unusual growth or a change in an existing mole. If the tool is trained on a light-skin patient data set, then it is designed to only function on light-toned skin, and therefore the AI-enabled solution leads to racial discrimination, bringing down its adoption among physicians and increasing legal risk for healthcare institutions.

A leader in IT services and consulting, TrustVector has hands-on experience in the AI development lifecycle and navigating unintended consequences. It emphasizes setting up AI governance to ensure the integrity of solutions as a business imperative. “By establishing robust AI governance, we help business leaders and innovators rethink how applications are built, delivered, and used,” says Sonja Petrović, co-founder and Chief Scientific Officer (CSO) of TrustVector.

TrustVector works across the entire AI solution lifecycle. Risk assessment and analyses are multidisciplinary: technical, legal, and ethical components are incorporated to provide a comprehensive understanding of AI solution’s trustworthiness.

To this end, TrustVector uses a framework that delivers a multidimensional vector across various AI trustworthiness parameters, such as: accountability, transparency, fairness (bias), and privacy. TrustVector’s multifaceted approach enables AI developers, users, solution buyers, investors, and regulators to proactively uncover issues, mitigate them and reduce the risk of unintended consequences caused by AI solutions.

TrustVector ties in insights from experts having diverse industry backgrounds. It offers a service, Fractional Chief AI Trustworthiness Officer, to assist organizations in instilling internal standards that help establish, operationalize, and monitor trustworthiness throughout the solution development lifecycle.

Their multi-disciplinary teams include experienced industry experts, seasoned C-suite advisors, academics (statisticians, computer scientists, data scientists, and ethicists), and digital transformation veterans.

Assessment and analysis outcomes provide actionable insights, guiding the set up of policies and procedures that ensure trust parameters are inherent from ideation to production. From design to implementation, TrustVector’s teams partner closely with their clients to provide guidance and offer recommendations covering AI governance, solution improvements, and process enhancements. The team also tailors their client’s AI strategies and frameworks to monitor the progress towards AI trustworthiness.

By establishing a trustworthy AI governance, we help business leaders and innovators rethink how applications are built, delivered, and used

The primary benefit of TrustVector’s approach is increased transparency across the entire AI lifecycle, giving organizations a clear current state picture, and actionable insights to advance their AI trustworthiness maturity.

TrustVector offers an external perspective and independent third-party assessments aimed at “advancing trust between humans and machines”. AI leaders can demonstrate their solution's reliability to potential buyers and regulatory bodies and increase user adoption. The company’s time-tested approach also presents investors an excellent opportunity to vet potential groundbreaking AI innovations, which is crucial to inform critical business decisions.

“Our human-led, technology-enabled approach, combined with trustworthy AI governance and solution assessment framework, helps our clients untangle ethical AI knots at every stage of the AI solution lifecycle, and enables organizations to confidently navigate the AI ethical conundrum,” says Jevtic.