Since artificial intelligence tools such as ChatGPT appeared at the end of 2022, AI has quickly become a part of the daily lives of millions of people.
From searching for information, planning work to creating content, AI brings unprecedented convenience.
However, along with the benefits are the risks related to privacy, accuracy and how people depend on this technology.
Dr. Shruti Patil, Director of the Symbiosis Institute of Artificial Intelligence (India), emphasized that 2026 will be the time when users need to be more alert than ever when using AI.
According to her, AI can support humans, but cannot and should not replace professional decisions, especially in the fields of health or finance.
Dr. Patil believes that the first important thing is to understand what AI can do. Repeated tasks, needing to synthesize information from many sources, are areas where AI is clearly effective. For example, planning a trip, which may take a few days to learn, now only takes a few minutes with ChatGPT's support.
Similarly, tasks such as creating basic content, designing invitation cards or building simple applications can also be done quickly thanks to popular AI tools.
However, Ms. Patil warned users to absolutely not share sensitive information with AI. Data such as personal identification information, financial accounts or passwords need to be strictly protected.
Large language models are trained from huge amounts of data globally, so providing private information can pose unpredictable risks.
In important decisions, AI should only play a reference role. Users can ask AI about stock market trends or medical concepts to understand more clearly, but should not completely trust.
You can use AI to understand more about what a doctor says, but you cannot replace your doctor with AI," Dr. Patil emphasized.
Another issue she mentioned is the "AI hallucination" phenomenon, when the tool provides information that sounds reasonable but is actually misleading.
According to her, AI often works well with short, simple data, but is prone to difficulties when processing long or complex documents.
Therefore, for office work or important data, users should not rely on free versions but need to use paid tools and always check the results.
Dr. Patil also believes that cross-checking information provided by AI is mandatory, even for small tasks but with great impact. Currently, no AI tool ensures absolute accuracy and consistency, which is still a major weakness.
In addition, she expressed concern about the abuse of AI tools, especially in acts of infringing on women's privacy through fake images and videos.
According to her, the responsibility lies not only with users but also with technology companies and the government. Countries need to build strict AI policies, clearly stipulate the types of data allowed to be used and prohibited behaviors, and require AI businesses to strictly comply.
“AI is a powerful tool, but only really useful when used properly and controlled,” Dr. Patil concluded.
