The Dark Side of the Algorithm: Ethical and Legal Challenges of Using AI for Career Planning
Author: Anna Puhachova – Career Counsellor, Ambassador of the National Europass & Euroguidance Centres in Ukraine, Director of the NGO “European Social and Economic Institute”.
Artificial Intelligence has fundamentally changed access to career counselling, evolving into a virtual advisor that is available at any time of the day. The ability to get an analysis of one's experience or options for professional development 24/7, regardless of schedule or geolocation, creates a sense of maximum convenience. However, this continuous availability often undermines the issue of digital security, which can lead to the leak of personal data, the construction of false career paths, and even reputational or legal problems during employment.
What data should not be shared with the algorithm?
A specific feature of interacting with open language models is that all data entered by the user is stored on developers' servers and used for further system training.
It is critically important to avoid uploading identification data, government documents, educational and personnel documentation, original CVs, and sensitive personal information into the system. Therefore, before assigning AI the task of analysing a professional profile, the text must be anonymised: remove contact details, exact names of employers, as well as any sensitive information about health, financial or family status. It is advisable to describe the professional background as broadly as possible.
How to avoid algorithmic bias?
AI learns from large datasets that often contain historical and societal stereotypes. Because of this, the algorithm may unconsciously limit choices, proposing career tracks based on a person's age or gender, rather than their actual competencies.
To obtain an objective result, the algorithm must be consciously guided. It is recommended to add clear instructions to your prompts regarding the generation of career development options based solely on professional skills and interests, demanding the system to completely ignore demographic characteristics. This helps to bypass built-in discriminatory patterns.
Which stages of employment are safe to delegate to AI, and which are not?
Delegating the writing of cover letters or the completion of test assignments to AI has become a common practice. However, from a legal and ethical perspective, this creates significant threats, as employers increasingly use software to detect generated texts.
Submitting entirely machine-generated text as one's own work is a violation of professional ethics, which usually leads to a rejection in employment. It is advisable to use AI exclusively as a supplementary advisory tool: for generating ideas, creating structure of the document, or checking grammar. The content must be written by a person himself/herself.
How to protect yourself from false information and legal risks?
AI systems are prone to generating convincing but completely false information (so-called "hallucinations"). There is a high risk of receiving recommendations regarding non-existent grants, links to repealed employment laws, or advice on taking courses that do not provide legitimate qualifications.
The responsibility for any career steps lies exclusively with the user, so all information generated by the algorithm must be carefully verified through primary sources: official government portals, official websites of grant providers, relevant funds or institutions, etc.
Conclusion
AI is not a substitute for a professional, but a powerful navigator that needs to be consciously steered. The use of AI in career planning significantly accelerates analytics and the generation of new ideas, but only under the condition of strict adherence to digital hygiene and critical thinking. The best strategy is to trust algorithms with routine data processing while retaining control over your own privacy, mandatory fact verification, and final decision-making. Ultimately, the responsibility for a successful and safe professional future always remains with the human.
