TCS Responsible AI Assessment for Generative AI

The surge in Generative AI (Gen AI) usage has resulted in rise of AI solutions like coding assistants, conversational assistants, learning assistants, operations assistants etc... which help workers, agents, and professionals to be more productive.

This technological leap has given rise to new category of issues like disclosure of confidential information, unethical manipulation, algorithmic bias, infringement of intellectual property rights, socioeconomic Inequality, and many more. This can lead to ethical concerns, and severe legal issues with high penalties for an organization. In addition, success measure for an AI system is to empower humans and improve sustainability -Imagine the risk in a scenario with conflicting objectives of human and AI assisted systems!

To reduce the associated risk, it is important to make these AI solutions secure, responsible and avoid any potential legal, security or ethical threats to humanity.

TCS offers Responsible AI (RAI) Assessment service to our customers and partners in gauging their Azure Open AI based Gen AI solutions against the core tenets of RAI: Security & Reliability, Accountability, Fairness & Ethical, Transparency and Explainability, Identity and Privacy Protection. This RAI assessment involves assessment of an Enterprise AI implementation across five important dimensions – Business, Process, Governance, Technical and People.

• Business Alignment – covers Business Objective and strategy for RAI at the enterprise. • Process Alignment – covers the existence of various processes for RAI usage at the enterprise level. • Governance Alignment – covers the governing mechanisms in place for Gen AI usage and implementation. • Technical Alignment – covers technology, tools, and guidance available for responsible Gen AI implementation. • People alignment – covers awareness, training, and competency building of workforce as per their roles.

Outcome of assessment is an insightful and actionable RAI Assessment report which will include - RAI maturity Score card, Gaps and Risks with Priority, and Recommendations with proposed Roadmap. This will assist organizations to gauge their current state of responsible use, devote directed efforts and reach desired level of maturity in swift and coherent manner.