ChatGPT Travel Planning: Experiences with Generative AI

Lisa Ernst · 02.12.2025 · Technology · 8 min

Artificial intelligence (AI) is increasingly shaping our daily lives, from travel planning to university admissions. This development raises questions about fairness, bias, and transparency.

Source: Own illustration

AI in Everyday Life

In everyday life, AI is increasingly used for complex decisions. Many people open a chat window for travel planning, formulation assistance, or medical background research. Travel portals, universities, and care organizations are integrating generative AI into their core processes. Phocuswright shows that over half of US travelers have already tried generative AI, and about a third are using it specifically for travel planning ( phocuswright.com, phocuswright.com). US universities like Virginia Tech and the University of North Carolina are using AI systems to pre-evaluate application essays and transcripts ( AP News, GradPilot). The US Department of Health and Human Services (HHS) is launching a competition with up to $2 million in prize money for AI solutions to relieve caregivers and people with disabilities ( hhs.gov, acl.gov).

Areas of Application

The application of AI extends across various areas:

Travel Planning with ChatGPT

Generative AI assistants enable more detailed travel planning. Instead of simple search queries, users can formulate complex requirements, e.g., "We are traveling with two children, need family-friendly hotels, want a maximum of two hotel changes, and value the most direct connections possible." Phocuswright reports that more than half of US travelers have tried generative AI and about a third are using it for inspiration, route planning, or restaurant recommendations ( phocuswright.com). Another analysis speaks of almost 40% of US travelers using GenAI tools for planning and booking ( phocuswright.com). Adobe saw a 3,500% increase in generative AI traffic on US travel websites within a year. Almost 29% of respondents use AI services for travel planning, and 88% of them report improved booking and travel experiences ( Adobe für Unternehmen). Emarketer shows that the proportion of consumers using generative AI for travel planning has increased from 8% to 24% in three years, with forecasts of 65% by the end of next year ( EMARKETER).

Source: urlaubschecker.at

ChatGPT and AI are revolutionizing travel planning.

A typical scenario is planning a one-week trip to Portugal, where ChatGPT considers budget, travel dates, children's ages, and interests to generate route suggestions with base locations, excursions, and restaurant examples. These suggestions are then reviewed and adjusted on booking portals. Travel blogs and tools offer guides on using ChatGPT as a "Trip Planner" with prompt examples ( Unstumbled, traveldifferently.org, skywork.ai). YouTube offers visual examples of travel planning with ChatGPT.

Travel providers are also upgrading. Booking Holdings points to AI-powered services such as Booking.com's "AI Trip Planner" and Priceline's AI assistant "Penny" ( Investors). Airbnb plans to become an "AI-first" app where agents can book entire trips for users ( Business Insider). However, Global Rescue warns that despite 22% of travelers using AI tools, trust and accuracy of information remain critical. International travelers particularly use AI to overcome language barriers and with visa questions ( globalrescue.com). Travel planning with ChatGPT provides initial ideas but does not replace subsequent verification on booking portals, with airlines, or official tourism sites to catch errors such as outdated opening hours or misleading visa regulations.

Source: urlaubschecker.at

Detailed travel plans like this can be created with ChatGPT in a very short time.

AI in University Admissions

While applicants in the US are often warned against having ChatGPT write their college essays, universities like Virginia Tech and the University of North Carolina are using AI systems to pre-evaluate essays ( AP News, GradPilot). Forbes describes these systems as a hybrid of human and machine, where AI provides an initial assessment and humans make the final decision ( Forbes). Some law schools, such as the University of Miami and the University of Michigan, have introduced optional essays where applicants are explicitly encouraged to use generative AI ( Reuters).

However, many universities are skeptical about applicants’ use of AI. A survey by Kaplan shows that only a small percentage of colleges explicitly allow generative AI in application essays, while a larger percentage prohibit it or are against it ( kaplan.com). Advisory sites like Spark Admissions and Top Tier Admissions warn against having AI write essays, as the texts appear stylistically interchangeable and are recognizable to admissions professionals ( Spark Admissions, toptieradmissions.com). A typical scenario is that an applicant writes their essay themselves and uses an AI tool for structural suggestions and phrasing variations, while the university uses another AI system for pre-screening the text. Two AI systems interact indirectly, without either side knowing the workings of the other model.

AI-powered systems are also used for predictions of academic success and scholarship allocation. A study commissioned by the American Educational Research Association shows that predictive models trained on historical performance data can systematically lead to disadvantages for ethnic minorities when used for admissions or funding ( aera.net).

AI Solutions for Healthcare

In the healthcare sector, the focus is on relieving the burden on informal caregivers and professional healthcare staff. The US Department of Health and Human Services (HHS) has announced the "Caregiver Artificial Intelligence Prize Competition" with up to $2 million in prize money for AI solutions ( hhs.gov). The Administration for Community Living (ACL) promotes AI tools that support safe, person-centered care at home and relieve employers in planning, training, and staffing of healthcare personnel ( acl.gov). Industry portals such as LeadingAge and eWeek emphasize concrete use cases such as intelligent appointment and medication management, early detection of health risks, and reduction of documentation burden ( LeadingAge, eweek.com).

Research shows that AI can support informal caregivers in decision-making, information seeking, and stress management ( PMC). One example is "ADQueryAid," a language model-based system that provides contextual information to relatives of people with dementia ( Nature). Companies are already offering AI solutions, including fall detection, smart environment recognition, chatbots for social interaction, and tools for coordinating appointments and documents ( newdays.ai, theflowspace.com). In practice, a daughter caring for her parent might receive reminders via an app for medication times, alerts about unusual movement patterns of her father, and suggestions for medical consultations. A chatbot offers emotional support and links to counseling services. The HHS competition aims to promote such practical solutions. YouTube videos about the "AI Caregiver Challenge" show how AI automates routine tasks and detects risks early without replacing human connection in care.

Source: insights.daffodilsw.com

Interacting with ChatGPT makes travel planning more efficient and personalized.

Opportunities & Risks

The opportunities of AI lie in reducing friction, faster orientation, and better support for overloaded systems. The risks include non-transparent evaluation logics, hidden discrimination, and a gradual habituation to the assumption that AI always knows the best solution.

Ethical Aspects

When AI helps decide on university placements or care support, ethical questions come to the fore. In 2021, UNESCO adopted the "Recommendation on the Ethics of Artificial Intelligence," which establishes transparency, fairness, protection of human rights, and human oversight as central principles ( UNESCO). In 2025, UNESCO published specific guidelines for the use of generative AI in education and research, concerning age limits, data protection, and the design of examination formats ( UNESCO).

Analyses show that algorithmic systems in education can reinforce structural inequalities when trained on biased historical data. The OECD shows in "Battling Algorithmic Bias in Education" that this can lead to systematically worse success chances for students from certain groups ( OECD). An AERA study concludes that predictive models for forecasting academic success can amplify racial disparities, even if "race" is not explicitly used as a variable ( aera.net). Bias in educational AI manifests in automated essay scoring, course recommendation systems, or the allocation of support services ( Schiller International University, yipinstitute.org, rene.kizilcec.com). A review article on fairness in Educational AI summarizes that AI systems in education tend to reproduce existing inequalities, and that technical "debiasing" approaches have limited effect without institutional changes ( arXiv).

A study from the University of Washington shows that people working with a slightly biased AI model adopt its distortions instead of correcting them. In the context of AI-assisted personnel selection, the model's racial biases were reflected in the test subjects' decisions ( The Washington Post). This means that in university admissions, when AI evaluates application essays and humans predominantly adopt this evaluation without systematic counter-checks, bias can be quickly scaled. Transparent disclosure, independent audits, the involvement of affected groups, and clear rules for where AI functions as a support rather than a decision-maker are becoming central governance issues. Organizations like UNESCO and the World Economic Forum emphasize that AI systems in schools and universities should always be used with clearly defined responsibilities, data protection rules, and understandable decision-making processes ( UNESCO, weforum.org).

Practical Implications

In everyday life, the role of AI is shifting from a toy to a silent co-decider. In travel planning, ChatGPT supplements traditional search engines with dialogical advice and personalized route suggestions. In university admissions, AI systems read essays and transcripts before a human opens the file. In healthcare, AI solutions are emerging that help decide when relatives can be relieved or when a doctor's visit is necessary.

Practically, this means:

• AI is excellent as a first instance for structuring, e.g., in travel planning or drafting an application essay. However, the final decision should consciously remain with the human.
• Institutions using AI in admissions or care should disclose where and how they do so and allow for independent audits.
• Fairness and accessibility must be considered design criteria from the outset, not a subsequent compliance add-on.

AI in everyday life is here to stay. The crucial question is whether we will design it as a covert judge or a transparent assistant.