Why Rule-Following AI Matters for Your Business
At Ethical Tech Matters, we are dedicated to helping organizations implement effective and ethical AI governance strategies. Our founder, Marian Newsome—an AI ethics consultant and IEEE CertifAIEd Lead Assessor—recently conducted an eye-opening experiment that underscores why responsible AI governance is not just a regulatory formality but a strategic imperative for businesses today.
The Experiment – AI Meets the Quartiles Puzzle
Marian took a moment during the Thanksgiving holiday to explore the capabilities of three advanced AI models—ChatGPT, Claude, and Gemini—through a simple yet revealing word game called the Quartiles puzzle. The rules of the game were clear:
- Combine letter tiles to form words.
- Use all letters on each tile in their given order.
- No splitting or rearranging allowed.
The surprising outcome? All three AI systems struggled to consistently follow these basic rules, often providing incorrect answers with alarming levels of confidence. This small experiment revealed an essential truth about the limitations of today’s AI systems—especially when they face tasks requiring strict rule adherence.
Why Even Simple Rules Challenge AI
The findings shed light on three core challenges in AI systems:
- Natural Language Understanding – AI often fails to accurately interpret nuanced or complex rules.
- Contextual Awareness – Current models lack the ability to recognize constraints within specific contexts.
- Abstract Reasoning – The generalization required to consistently follow rules remains a developing capability.
While this experiment might seem trivial, the same challenges can have serious real-world consequences when AI systems are deployed in industries such as financial services, healthcare, and manufacturing.
Real Impact Across Industries
Marian’s findings have direct implications for businesses that rely on AI in critical, high-stakes environments:
- Financial Services face substantial compliance risks when AI systems confidently deliver non-compliant outcomes, like incorrect loan approvals.
- Healthcare risks arise when AI systems misinterpret clinical guidelines, creating significant patient care issues.
- Manufacturing experiences challenges when quality-control AI systems fail to detect errors, necessitating manual intervention to maintain safety standards.
Behind these failures lies a common issue—AI systems displaying high confidence in outcomes that are fundamentally flawed, which can lead to costly errors.
The Role of Governance Frameworks
At Ethical Tech Matters, we believe that the solution to these challenges lies in rigorous governance. Marian emphasizes the importance of adopting trusted AI governance frameworks to mitigate risks and ensure responsible AI implementation. Essential tools include:
- IEEE P2863™ Standard – Focused on organizational governance with principles of transparency, accountability, and fairness.
- NIST AI Risk Management Framework (AI RMF 1.0) – Designed to address trustworthiness by mapping, measuring, and managing risks.
- EU AI Act – A proposed comprehensive regulation that enforces a risk-based approach to AI governance.
These frameworks provide businesses with essential guidelines to evaluate, monitor, and improve their AI systems while aligning with ethical practices.
A Real-World Case Study
Marian also shared an example of a leading financial institution that deployed AI to speed up loan decisions. Despite being programmed to follow strict regulatory rules, the system routinely approved ineligible applications, leading to compliance violations and fines. The root cause? The AI incorrectly generalized patterns from its training data, prioritizing speed over accuracy.
Through the implementation of governance frameworks like NIST AI RMF 1.0, along with ongoing audits and human oversight, the bank was able to rebuild trust in their AI system, ensuring compliance and improved effectiveness.
Why Governance is a Strategic Imperative
This experiment—conducted in a kitchen, of all places—highlights a critical message for enterprises. Effective AI governance is not just about avoiding regulatory repercussions; it is a tool for empowering organizations to make AI systems reliable, trustworthy, and aligned with their goals.
Partner with Ethical Tech Matters
At Ethical Tech Matters, we are committed to driving positive change in technology by promoting ethical, well-governed AI systems. Take these next steps to strengthen your organization’s approach:
- Subscribe for Updates – Stay informed about AI governance insights and best practices from Ethical Tech Matters.
- Schedule a Consultation – Partner with our expert team for tailored solutions to your AI governance challenges.
- Download Our Framework Assessment Tool – Begin identifying and addressing governance gaps within your organization using a comprehensive assessment aligned with IEEE P2863™, NIST AI RMF 1.0, and the EU AI Act.
Don’t leave your AI systems unchecked. Work with Ethical Tech Matters to build reliable, ethical, and high-performing AI for your business.
To further explore the challenges of AI systems and the pressing need for effective governance, check out Marian Newsome’s insightful article, When AI Can’t Follow Simple Rules: A Critical Warning for Enterprise Leaders, on Medium. This detailed piece dives deeper into real-world examples of AI’s limitations and underscores why ethical technology practices are essential for businesses today. It’s a valuable read for anyone committed to building reliable and responsible AI systems.
Marian Newsome’s compelling insights remind us that even seemingly small failures in AI rule-following can carry significant consequences. At Ethical Tech Matters, we’re here to help your organization navigate these challenges and lead with integrity in the age of AI. Take action today and equip your business for success.
