ISE’s Student Recruitment Survey provides insight to the extent of AI adoption in early careers recruitment, its perceived benefits, and the challenges associated with its use in hiring.
AI is increasingly influencing the world of work, particularly within recruitment and selection. Its integration is transforming traditional methods by introducing efficiencies in candidate screening, reducing bias, and supporting data-driven decision-making. At the same time, concerns remain around transparency, fairness, and the ethical implications of AI-driven tools.
This year’s ISE Student Recruitment Survey took a deep dive into AI. In addition to exploring employers’ use of AI in recruitment and their perspectives on its benefits and drawbacks, we also examined whether they permit candidates to use AI during the hiring process.
What’s the extent of AI adoption in selection processes?
Compared to last year (23%), a greater proportion of employers (33%) reported that generative AI was leading them to redesign their selection processes, while a further 46% said they were currently reviewing their processes in response to generative AI.
In this year’s survey, we have explored automation and AI in recruitment. Automation refers to using systems with preset benchmarks or rules to perform specific routine tasks or make automatic decisions about whether to reject or progress a candidate.
AI refers to using systems which scan, rank or filter more complex information to make decisions or recommendations by learning from data such as identifying the best candidates or analysing video interview responses.
While automation and AI are increasingly present in recruitment, most high-touch, interpersonal activities remain predominantly human-led.
Traditional processes such as in-person interviews, group tasks, role play, and case studies are conducted almost entirely by humans (94–100% fully human), indicating that organisations still value direct human judgement for activities requiring nuanced interpersonal evaluation.
Conversely, assessments that rely on clear scoring criteria - such as numerical reasoning (57% fully automated), situational judgement (53%), and verbal reasoning tests (49%) - are the most likely to be automated.
This pattern suggests automation is primarily being applied to structured, rule-based evaluations rather than subjective assessments.
The use of AI-only systems remains limited, with notable presence in game-based assessments (15%), and other psychometric tests (8%).
Hybrid human-system approaches are most common in strengths-based aptitude tests (42%), online job simulations (33%), and behavioural or values assessments (30%), pointing towards a transitional phase where technology supports but does not replace human decision-making.
Overall, the findings highlight a recruitment landscape where automation increasingly provides efficiency in objective tests, but AI adoption is still emerging, and human oversight remains critical, particularly in stages where candidate interaction, soft skills, and cultural fit are evaluated.
However, looking to the future, more than half of employers (62%) expect to use AI in their recruitment processes, and 70% anticipate increasing their use of automation in selection over the next five years. This indicates that AI is set to play a growing role in early careers recruitment.
What are the advantages and disadvantages of using AI in recruitment?
Employers also shared their insight about the benefits and drawbacks of using AI in their hiring process.
Most (94%) employers reported that using AI increased speed and efficiency in the recruitment process and increased ability to analyse large volumes of data and identify patterns and trends (81%). The latter shows a notable increase compared with last year, rising from 58% to 81% — a 23-percentage point gain. This suggests employers are increasingly recognising AI’s potential to help manage the growing volume of applications.
Perceptions of AI’s costs and benefits have also shifted. A quarter (25%) of employers now see AI as cheaper than using people to perform equivalent tasks, up from just 10% last year — a 15-percentage point increase. While still a minority view, this indicates a growing awareness of AI’s potential to reduce costs.
Fewer employers reported using a recruitment process outsourcing (RPO) this year. It may be that some organisations are beginning to explore whether automation and AI can replace or supplement outsourcing, though this is a trend that will need to be tracked in future surveys.
For the disadvantages of using AI, 71% of employers reported that they preferred a more human-centric approach in the recruitment process and worried about the potential for bias in AI decision-making (67%).
You can read more in-depth data on the benefits and drawbacks of AI in recruitment in our Recruitment Survey.
Do employers want candidates to use AI in the recruitment process?
This year, we also asked employers about their views on candidates using AI in the recruitment process.
Almost half (46%) reported that they have no problem with candidates using AI at some stages of the process. A further 29% recommended that candidates do not use AI in the recruitment process but did not take any action if candidates used it.
However, 45% of employers said they had not provided applicants with any guidance on when it was or was not appropriate to use AI. Among employers who permitted AI use, the most common applications were in writing cover letters or CVs (61%), followed by completing online application questions (41%).
Read how to guide candidate use of AI and Shoosmiths shares their experience of using AI in recruitment.
To what extent are candidates using AI to ‘cheat’ recruitment?
As the recruitment landscape evolves with the integration of AI and other digital tools, employers are placing increasing emphasis on fairness and transparency in selection processes. However, the use of AI by candidates during assessments has introduced new challenges in maintaining integrity.
Compared with last year (22%), only 15% of employers in 2025 reported that they had never suspected or identified cheating in assessments, indicating that suspected misconduct is becoming more common. Furthermore, the proportion of employers who said they frequently encountered cheating more than doubled, rising from 7% last year to 15% this year.
The most commonly reported form of misconduct was candidates using AI during interviews without disclosure or permission, cited by 61% of respondents.
While most employers have strategies to detect or address misconduct, 30% reported they did not have any approach to do this. Among those with measures, the most common were the creation of cheating prevention guidelines and policies (33%) and the use of time limits on assessments (29%).
The development of and increased access to AI is having major implications for the student recruitment market. This is likely to continue to be a hot topic while both candidates and recruiters explore the possibilities and drawbacks of AI further.
More detailed analysis and data on AI in early careers recruitment can be found in ISE Student Recruitment Survey.