TLDR: Globant, a digitally native company, is championing a new standard for ethical AI, advocating for AI bias to be as rigorously measured and managed as system uptime. This push emphasizes the critical need for quantifiable metrics and continuous auditing to ensure fairness, transparency, and accountability in artificial intelligence systems.
Buenos Aires, Argentina – In a significant move towards fostering more responsible artificial intelligence, Globant, a global leader in digital transformation, has called for AI bias to be made as measurable and trackable as traditional system uptime. This initiative underscores a growing industry demand for concrete, quantifiable metrics in the realm of AI ethics, moving beyond abstract principles to actionable governance.
Globant’s stance is deeply rooted in its comprehensive AI Manifesto, a set of guiding principles designed to ensure the ethical development and deployment of AI. The manifesto explicitly promotes data-driven outcomes that are unbiased across various demographics, including race, ethnicity, gender, and socio-economic status. A core tenet is the commitment to algorithmic accountability, ensuring that AI products are transparent in their purposes and results, and that risks of user data misuse are preempted.
‘We are responsible for the data and how its use may affect society,’ states Globant’s AI Manifesto, emphasizing the imperative to ‘avoid biased results as much as possible to eliminate unfair outcomes.’ This commitment extends to promoting ‘respectful data’ practices, ensuring data cleanliness, accuracy, and compliance with privacy regulations, alongside strict reliability and security standards.
To operationalize these principles, Globant has invested heavily in internal capabilities and external partnerships. All of Globant’s multidisciplinary PODs (teams) undergo extensive AI training, equipping them with the expertise to identify AI adoption opportunities while managing associated data and security risks. Every POD subscribes to the AI Manifesto, integrating ethical considerations into every project. The company has also actively funded startups dedicated to tackling AI bias, further demonstrating its commitment to mitigating negative technological impacts.
Roberto Contreras, a key figure at Globant, has highlighted the importance of continuous auditing for AI agents. He stresses that monitoring should not be a one-time validation but an ongoing process, utilizing dashboards, logs, bias reviews, and performance tracking. ‘Continuous auditing ensures that your AI agents are monitored in real-time,’ Contreras noted, advocating for visibility into agent decisions and key performance indicators (KPIs) to enable traceability and post-hoc analysis. He also underscored the critical role of ‘human in the loop’ mechanisms, ensuring human oversight and the ability to intervene in novel or high-risk situations, even with autonomous agents.
Also Read:
- Globant Unveils AI Pods: A Paradigm Shift in IT Services with Outcome-Based Subscription Model
- Microsoft’s AI-Driven Efficiency Yields Over $500 Million in Savings, Highlighting Ethical AI’s Business Value
This call for measurable AI bias aligns with broader industry discussions on AI governance, particularly those taking place at events like the AI for Good Global Summit. Such forums aim to bridge the gap between high-level ethical principles and practical, on-the-ground regulatory implementation, seeking pathways to effective and trustworthy AI governance at speed. Globant’s push for quantifiable metrics is a significant step towards making AI ethics a tangible and auditable component of AI development, akin to the established practices of measuring system performance and reliability.
