Personality assessments are widely used in applied psychology, particularly in personnel selection, where they help predict workplace behavior. Their validity, however, critically depends on honest responding. Faking—intentionally distorting responses to create a favorable impression—has been a long-standing challenge. With the rise of generative artificial intelligence (AI), this challenge has reached a new level: applicants now have access to powerful tools capable of producing highly tailored, convincing responses that mimic desirable personality profiles. What was once a matter of individual strategy and test-taking skill has become a qualitatively different challenge that raises fundamental questions about the integrity of assessment results. This presentation examines how AI is transforming faking in personality assessment, the limitations of current prevention and detection methods, and possible strategies to safeguard validity. We synthesize findings from recent empirical studies comparing AI-generated responses with human faking attempts, review the effectiveness of traditional detection tools, and explore emerging approaches such as behavioral data analysis, test design adaptations, and preventive interventions. Evidence suggests that AI can, under certain conditions, outperform human fakers by producing highly socially desirable profiles aligned with job requirements. Traditional detection methods often fail to reliably identify such responses. At the same time, test formats such as forced-choice and situational judgment tests show greater resistance. Preventive strategies—including warnings, proctored testing, and technical safeguards—offer practical ways to reduce faking. Importantly, AI use may not always be a threat: in some contexts, the ability to leverage AI tools could itself represent a job-relevant skill. Generative AI therefore presents both risks and opportunities for applied psychology. Safeguarding the integrity of personality assessment requires a shift from purely detection-focused approaches toward prevention and innovative test design. Ongoing research and collaboration are crucial to ensure assessments remain valid, fair, and practically useful in the age of AI.