Ai Literacy without Dependency

What is Automation Bias?

Automation bias is when you believe a computer or AI just because it gave an answer, even when the answer is wrong.

Imagine a calculator says that 2 plus 4 equals 5. If you say, “The calculator said it, so it must be true,” without checking, that is automation bias.

NIST describes automation bias as giving too much trust or deference to automated systems, which can make people rely on AI too much.

What is Confabulation?

Confabulation is when an AI gives information that sounds confident and convincing but is wrong, false, or made up. It may invent facts, names, sources, quotes, or explanations.

Example: An AI may confidently recommend a supplier, cite a news article that does not exist, or give a false market price.

Remember: AI can help you, but always verify important information before you act on it.

Reference: National Institute of Standards and Technology. Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile (NIST AI 600 1, July 2024), section 2.2

AI literacy is not only about using AI. It is also about understanding its hidden costs, including bias, digital inequality, underrepresentation, energy use, water use, e-waste, mineral extraction, and the concentration of AI power in a few countries.

UNESCO warns that AI can embed bias, deepen inequalities and digital divides and underrepresent certain cultures and communities if ethics and literacy are ignored.

UNU INWEH’s 2026 report estimates that by 2030; data centres could use about 945 TWh of electricity, nearly 3 percent of global use, water use could match the basic annual needs of roughly 1.3 billion people in Sub Saharan Africa and land use could exceed 14,500 square kilometres.

AI infrastructure could generate up to 2.5 million tonnes of e-waste per year. Critical minerals for AI hardware are often extracted in regions with weak environmental protection, leading to pollution, informal e-waste and health risks.

Most AI specialised computing capacity is concentrated in a small number of countries, especially the United States and China.
People in our focus regions are rarely told any of this. They see AI outputs, but not the infrastructure, data choices or environmental costs behind them.