AI is seeing widespread use across many industries, but its reliability still leaves something to be desired.
In the year 2026, AI is everywhere. Schools, online journals, laboratories, and an ever-growing number of private companies are using AI systems for a wide variety of tasks, often in the name of speed and efficiency.
As anyone who has used an AI chatbot before can tell you, however, that speed can come at a real price.
Hallucination AI problem
If you were to tell a chatbot that “you’d be willing to pay millions of dollars to have a pizza right now,” the LLM powering that chatbot might interpret that statement literally; rather than reading your statement as a ridiculously exaggerated desire for pizza, it might consider that you would actually be willing to pay one million male deer for a slice of pizza.
This kind of misinterpretation, or “illusion” in the context of AI, is harmless when you’re asking a chatbot silly questions. Less dangerous is when AI provides inaccurate or even completely fabricated information for a pharmaceutical company testing drug interactions, or a supply chain manager trying to predict the best shipping routes during a period of political turmoil.
AI hallucinations might be more manageable if it were always obvious when a program is creating or distorting information, but because most LLMs are designed to sound confident and agreeable, it can be difficult to tell truth from lies without a thorough fact-check. Indeed, a fact-checking program intended to fact-check its users is counterproductive and, frankly, self-defeating.
There are two very important reasons for seeing things under AI: one, most LLMs are not programmed to inform users if they don’t know something, and two, most training data that LLMs learn from is itself full of inaccuracies and assumptions.
These combined factors make it easy for LLMs to not only be wrong but to be sure, and when businesses all build their models with the assumption that AI almost always gives clear, correct, and accurate answers, those businesses and the people who work with them can end up sometimes basing their actions on false information.
This is not to say that businesses are unaware of the fact that AI can and does hallucinate; However, some companies have started going out of their way to develop AI models that directly address the causes of AI hallucinations.
Improving AI Credibility in the Real World: A Case Study
One such company working to improve the reliability of AI is Vertus, an AI company based in the Isle of Man. Its founders, Julius Franck, Alex Foster, and Michal Prywata, have created a reasoning system designed to recognize when certain patterns work and when they don’t, thus helping them avoid making the same kinds of assumptions that many other LLMs may continue under similar circumstances.
To test their AI, Vertus had a system trade in financial markets during 2025. During this period, the company reported good results.
Vertus attributes its success to its system’s ability to quickly adapt to new market patterns. To do this, the AI is designed to ask if a given pattern still works in a given situation. If it doesn’t, the system recognizes the fluctuation, stops, and rebuilds its thinking about what’s really going on.
As a kind of failsafe, the AI is also designed to tell its user when it can’t come up with an answer to a question, reducing its chances of coming up with a sure but illogical answer.
As Vertus’ tests have come back positive, the company has begun extending its AI solutions to healthcare, scientific research, and supply chain management.
While Vertus isn’t the only organization working to improve the reliability of AI, its success serves as a useful indicator of how, so far, it’s proven to be important. Building AI systems to test new knowledge against what they already know and tell users when they don’t know something are important first steps in reducing the perception of AI, though whether those systems will become commonplace anytime soon remains to be seen.
There is still work to be done
Even in the few years since chatbots became popular with the launch of ChatGPT in 2022, the practical and theoretical use of AI has grown exponentially. While that rapid growth has helped many businesses improve their bottom line by cutting costs, such rapid growth in a short period of time has its consequences.
AI illusions are still a serious problem, and as AI gains more traction in medicine, finance, education, and many other important industries, the need to address its tendency to provide quick, confident answers at the expense of reliability will become even more important.
AI’s ability to collect, organize, and analyze vast amounts of information in a matter of seconds could prove useful to organizations for years to come, but any progress must be linked to efforts to improve AI’s credibility before people consider expanding on its flawed foundations.
The information provided in this article is for general informational and educational purposes only. It is not intended as legal, financial, medical, or professional advice. Readers should not rely solely on the content of this article and are encouraged to seek professional advice tailored to their specific circumstances. We exclude liability for any loss or damage arising directly or indirectly from the use of, or reliance on, the information presented.
Digital Trends partners and external contributors. All contributor content is reviewed by Digital Trends editorial staff.
