Why Does AI Struggle to Tell Time?

In 2026, artificial intelligence can write code, generate realistic images, compose music, and hold complex conversations — yet it still struggles with something every 6-year-old masters effortlessly: reading a clock. This puzzling gap between AI’s extraordinary capabilities and its surprising blind spots reveals something fundamental about how these systems actually work. If you rely on AI for scheduling, automation, or time-sensitive business tasks, understanding this limitation is critical.

In a landmark study published on arXiv, researchers from the University of Edinburgh tested 7 major Large Language Models (LLMs) on their ability to interpret time from visual inputs — clocks and calendars. The results were sobering and carry major implications for anyone deploying AI in real-world applications in 2026.

Their tests spanned various visual formats: analog clocks with Roman numerals, clocks without second hands, digital displays, and calendar grids spanning multiple years. Despite these models being trained on billions of images and texts, their performance on temporal visual reasoning was far below what we’d expect from humans — or even from children.

The researchers wrote in the study: “The ability to interpret and infer time from visual inputs is crucial for many real-world applications, from scheduling events to autonomous systems. Despite advances in multimodal large language models, most research has focused on object detection, image labeling, and scene understanding, but hasn’t focused on temporal inference, which has left the time factor neglected in these systems.”

The research team tested models from various companies including OpenAI’s ChatGPT-4o, Google’s Gemini, Anthropic’s Claude, Meta’s Llama, the Chinese model Qwen 2 from Alibaba, and MiniCPM from ModelBest. They presented images with different clock styles — wall clocks, Roman numeral faces, clocks without second hands — as well as calendar images spanning the past 10 years.

How AI Models Were Tested on Clock Reading in 2026

In the clock test, researchers asked the large language models: “What time is shown in the attached image?” For the calendar test, they posed both simple questions like “What day does New Year’s Day fall on?” and complex queries such as “What is the 153rd day of the year?”

The researchers stated: “Reading clocks and understanding calendars requires complex cognitive steps, needing precise visual discrimination — to recognize the position of clock hands and calendar layout — as well as precise numerical thinking to calculate the number of days between two dates.”

This is not merely a curiosity — it’s a real-world constraint. Businesses using AI for meeting scheduling, automated reporting, or time-stamped workflows may encounter silent errors when AI misreads temporal visual data. The implications for autonomous systems, healthcare AI, and industrial automation are even more significant.

Surprising Results: Which AI Models Performed Best?

Overall, AI models did not achieve satisfactory results. They correctly read the clock time in less than 25% of cases and struggled particularly with clocks displaying Roman numerals or innovative hand designs. Even clocks lacking second hands caused significant errors. Researchers suggest the problem lies in detecting hand positions and interpreting angular relationships on the clock face — skills that require precise spatial reasoning that current vision models lack.

Notably, the Gemini model scored highest in the clock reading test, while ChatGPT-4o excelled in reading calendars and determining time with 80% accuracy. By contrast, most other large language models made errors in the calendar test at approximately 20%. This divergence is telling: models optimized for language reasoning (like GPT-4o) may handle structured calendar grids better, while purely vision-based tasks expose deeper limitations in spatial-temporal understanding.

Why This Matters for AI-Powered Business Tools in 2026

Rohit Saxena, one of the study’s authors and a PhD student at the School of Informatics at the University of Edinburgh, said: “Most people can tell time and use calendars from an early age, but our results show the significant gap in AI’s ability to perform what are considered very basic skills for humans. We should not overlook these problems if we want to integrate AI systems into time-sensitive real-world applications such as scheduling, automation, and assistive technology.”

He added, “Although AI can complete most of your homework, I wouldn’t recommend relying on it to meet any deadlines.”

As AI tools become more integrated into business workflows in 2026, understanding where they fall short is just as important as knowing where they excel. The time-reading limitation is a reminder that AI is a powerful assistant — not an infallible oracle. The key is knowing which tasks to trust it with, and which to double-check.

Want to get the most out of AI despite its limitations? The right prompts can dramatically improve AI accuracy on time-sensitive tasks. Learn how to craft better AI instructions with our free tools below.

🚀 Try our free AI Prompt Generator — craft prompts that get more accurate, time-aware responses from any AI model.

📰 Stay updated with the latest AI news — follow developments in AI capabilities, limitations, and breakthroughs as they happen.