Agentic AI: What is it? And does it pose a threat?
 |
| What is Agentic AI and is it safe? |
The tourism and travel industry is approaching a major turning point that could eliminate the need to spend hours comparing flight and hotel prices. Collaboration between technology companies and travel platforms is intensifying to develop new generations of artificial intelligence (AI) tools capable of handling all travel arrangements on behalf of users and at prices that suit their needs.
Currently, millions of travelers rely on AI tools to compare available flight, hotel, and car rental options. However, these tools are about to undergo a significant transformation with the advent of Agentic AI. Developers say that Agentic AI, an emerging type of AI, will be able to find and process bookings with minimal human intervention.
According to a report by The New York Times, reviewed by MobiAlo, travelers using Agentic AI will specify their travel criteria, such as dates and price range, and then provide their credit card information to the AI, which will automatically monitor prices and complete bookings for them.
Although these tools are still in their early stages, they are expected to grow rapidly. According to a September 2025 report by McKinsey & Company and Skift, a travel industry publication, 80 percent of travel executives plan to widely deploy Agentic AI tools within the next five years.
Currently, companies like Expedia, Google, Kayak, and Priceline are experimenting with these tools, with clear plans to bring them to market soon.
What is Agentic AI?
According to data reviewed by the website "mobi alo," Agentic AI, or "agent artificial intelligence," represents the next generation of AI tools capable of making decisions and completing tasks on behalf of humans with minimal human intervention. The capabilities of this type of AI are not limited to the travel and tourism industry.
For example, Agentic AI tools can manage procurement, book medical appointments, organize work schedules, and even complete small financial transactions or monitor resource consumption in companies—all automatically and intelligently.
Experts indicate that the widespread adoption of this technology in the travel sector is just the beginning. Other industries, such as retail, banking, and healthcare, will witness a similar wave of adoption in the coming years. Thanks to its ability to learn from data and adapt to market changes or user preferences, Agentic AI is expected to become an essential tool in the daily management of both individuals and organizations.
How does Agentic AI work?
Agentic AI is an advanced type of artificial intelligence that combines data analysis, decision-making, and continuous learning capabilities. It works by analyzing vast amounts of online information, including prices, availability, offers, and reviews, to identify the best options as quickly as possible.
Thanks to machine learning algorithms, Agentic AI can evaluate different alternatives and select the most suitable one, considering the risks, costs, and potential benefits of each decision. It also learns dynamically from past experiences, becoming more accurate and efficient over time and ensuring continuous performance improvement.
Furthermore, Agentic AI can connect to multiple services and platforms in real time, enabling it to manage a variety of tasks simultaneously.
A role that is not yet complete
Shilpa Ranganathan, senior product officer at Expedia Group, which develops Agentic AI-powered trip planning tools, says that the technology is still in its early stages and is currently only used as a tool to provide a more detailed level of information compared to older searches. She notes that Agentic AI tools are now able to collect prices for several hotels for the dates specified by the user, check availability, and then display this information in one place to facilitate comparison, leaving the final decision to the user.
The biggest obstacle
A report by Skift, "The State of Travel 2025," revealed that over 90 percent of travelers trust AI-generated travel information. However, a mere 2 percent are willing to grant these tools complete autonomy to make booking decisions or modify plans without human intervention.
Another report published by Booking.com in July showed that only 12 percent of respondents were completely comfortable entrusting AI with making travel decisions independently. Privacy remains the biggest obstacle, with more than half of those surveyed expressing concerns about how travel companies use their personal data.
This challenge presents a critical hurdle for developers if they hope to persuade travelers to provide their credit card information to AI, especially given that intelligent robots are still susceptible to hallucinations that could sometimes lead them to make irrational decisions.
Should we trust Agentic AI?
Oxford-certified AI consultant Hilda Malouf Melki told the website "mobi alo" that Agentic AI is not merely a comparison or assistance tool, but a step towards an intelligent agent that can alleviate the administrative burden on humans. However, she noted that trusting it requires developing rigorous security mechanisms to ensure the accuracy of the data it relies on. She pointed out that the role of this technology is not yet complete, as it is still in an experimental phase, paving the way for broader and more autonomous automation in the future.
Malouf Melki explained that the key advantage of Agentic AI is its limited autonomy, meaning it can operate according to predefined rules and standards and learn from past experiences to improve its decisions. This saves humans time and reduces errors resulting from work pressure or information overload, which is crucial in sectors like healthcare, where this technology can schedule appointments, monitor cases, and even suggest simple treatment plans in coordination with doctors.
Read more : 10 things you should never share with Chat GBT
Malouf Maliki adds that trust in Agentic AI is possible but conditional and not absolute at present. The risks associated with information hallucinations, privacy issues, and the possibility of software errors make it necessary to maintain human supervision during its use, especially when dealing with sensitive data or making important financial or medical decisions. She stressed that users need to ensure the accuracy and security of the data on which Agentic AI tools depend before allowing them to act with complete independence. They should be given limited independence in less dangerous tasks first, with reliance on them gradually increasing as they prove their accuracy and reliability.
Steps to building trust
Malouf Maliki explains that building trust in Agentic AI can only be achieved through the development of multi-layered verification systems that ensure the accuracy of the data on which the model is based before any step is taken, in addition to continuous auditing mechanisms that detect potential biases or hallucinations and prevent the transmission of false results to the user. Hence the importance of systems containing operating software that prevents the intelligent agent from making decisions that exceed its authority, with the necessity of having an immediate human intervention button that allows it to be stopped or its course modified when any unexpected behavior appears.
Is the era of manual research over?
For his part, Amer Dabbousi, the managing partner of Eagle Travel, a tourism and travel company, told mobi alo "Economy" website that there are limits to what artificial intelligence can do on its own. Manually searching for the best deals in the world of travel is still necessary in cases that require a personal assessment of quality or individual preferences, or when dealing with rapidly changing or incomplete information. This applies to the tourism and travel sector, he noted, pointing out that relying on human expertise remains necessary to ensure that choices meet the precise needs of users, and to evaluate qualitative factors such as hotel comfort or service level, aspects that are difficult for artificial intelligence to fully capture without human supervision.
Machines cannot replace humans
Daboussi emphasizes that reliance on artificial intelligence (AI) is gradually evolving to complement human expertise, creating a hybrid experience that combines the two. Agentic AI tools will make the process of searching for available options faster and more efficient, while humans will oversee and review the AI-proposed options to ensure they are free from bias and correct any potential errors before a final decision is made. We all know how serious the mistakes AI programs can make.
Daboussi believes that technology can accelerate some routine processes, but it cannot replace human expertise in assessing quality or making decisions based on qualitative criteria, whether in travel or other fields. He points out that granting AI technology excessive autonomy can be risky. Users must realize that these tools, however advanced they may seem, are still susceptible to bias, error, and mistakes that may be irreversible once damage has been done. Therefore, Agentic AI should be treated only as an aid, not as a thinking or decision-making body that replaces humans.
