Would you be comfortable knowing that AI had ranked your job application? What if using AI helped a colleague secure a pay rise, while you were still waiting for training?
Recent stories about AI in HR bring those questions into focus. October’s coverage has examined AI skills and salaries, automated CV screening, candidates receiving help during interviews, and employers trying to establish who is really applying for a job.
Each development presents opportunities, but also some uncomfortable decisions. How should employers reward AI skills? What makes an application trustworthy? And how much responsibility should recruitment software have when someone’s career is at stake?
Here are four stories worth understanding, along with supporting research and their practical implications.
1. AI skills are becoming part of pay and promotion decisions
On 14 October, Reuters reported on financial firms linking AI proficiency to pay and promotion.
A Selby Jennings survey of 500 finance executives found that nearly a quarter of firms connected raises, bonuses or promotions to employees’ AI capabilities. Another 30% required staff to disclose their AI use during performance reviews. The findings concern the surveyed firms, rather than a universal policy across financial services.
For employees, that makes the discussion about AI training rather more personal. It could influence how their contribution is assessed and which opportunities become available.
But what should an employer actually reward?
Someone might use AI every day and spend hours correcting its mistakes. Another employee might use it for a few carefully chosen tasks, save time and produce excellent work. Measuring usage alone would tell an incomplete story.
A useful assessment would consider accuracy, judgement, confidentiality and results. Can the employee spot an unreliable answer? Do they know when a task needs specialist review? Can they explain how AI improved the finished work?
Fair access matters too. Before making AI skills part of promotion decisions, employers should consider who has received training, which tools people can access and whether their roles offer comparable opportunities to use them.
Otherwise, an apparently modern performance measure could reward access as much as ability.
2. Can AI recruitment tools recognise a good candidate?
An October Financial Times column described a recruitment experiment at Addleshaw Goddard, comparing an AI agent’s assessment of shortlisted applications with human judgement.
The rankings differed. The agent favoured some polished applications containing claims that human reviewers questioned. People ultimately chose the successful candidate, who was not the AI’s highest-ranked applicant. This was one experiment, but it raises a useful question about what automated screening actually rewards.
There is supporting research behind that concern. A 2025 study published through the AAAI/ACM Conference on AI, Ethics, and Society found that, in controlled experiments, AI models preferred CVs they had generated themselves over human-written alternatives or those produced by other models, even when content quality was controlled.
The research describes a potential source of unfairness in AI recruitment. It does not establish that every commercial screening system behaves this way.
For applicants, the possibility is unsettling. Could their choice of writing tool affect their prospects, even when their experience stays the same?
For employers, it is a reason to test AI CV screening carefully. A ranking should come with evidence that relates to the job, rather than an unexplained score.
An impressive application has never guaranteed an impressive employee. Follow-up questions, work examples and appropriate practical exercises remain useful ways to examine what someone can actually do.
3. AI in job interviews needs clearer boundaries
The Institute of Student Employers’ Student Recruitment Report 2026, published on 14 October, highlights concerns about applicant conduct.
Some 92% of surveyed employers suspected or confirmed misconduct during assessment and selection. That covers misconduct generally; it does not mean 92% caught candidates cheating with AI. Unauthorised AI use during interviews and online assessments was the most common concern.
The report also found that 82% of employers provided guidance on acceptable AI use, up from 55% in 2025.
That guidance matters because “using AI” covers very different activities.
A candidate might use a chatbot to understand a job description, organise examples from their career or practise answering questions. Secretly receiving generated answers during a live assessment creates a different problem when the employer expects independent responses.
There is also a reasonable question about assessment design. If a role involves working with AI, employers could include an exercise where candidates are explicitly allowed to use it. They could then assess how well applicants check the output and explain their decisions.
A separate exercise could test knowledge without assistance. Both can have a purpose, provided the instructions are clear.
Candidates also deserve transparency from employers. If AI helps screen applications or assess interviews, explaining its role would make expectations less one-sided.
Earlier background from Greenhouse’s 2026 AI in Hiring Report explores candidate trust, verification and communication about AI use. It offers a broader context for the recruitment questions appearing in October’s news.
4. Fake candidates are bringing HR and cybersecurity closer together
A recent Financial Times report examined recruitment fraud and identity checks, including impersonation, chatbot-assisted deception and false identities used to gain access to company systems.
The reporting describes HR teams working more closely with cybersecurity specialists. It also questions what increasingly suspicious hiring processes mean for genuine applicants.
There is earlier supporting material from recruitment platform Greenhouse. Its September guide to detecting candidate fraud discusses synthetic identities, third-party stand-ins and AI-assisted interview scripting. It also stresses that a warning sign should prompt investigation rather than automatically disqualify someone. As a supplier of hiring software, Greenhouse has a commercial interest in this area.
The practical challenge is balancing verification with a fair experience.
Employers need confidence in someone’s identity, particularly where a role provides access to sensitive information. Applicants need to understand why checks are happening and how to resolve a mistake.
A structured process can help. Consistent questions, relevant evidence and a clear route for reviewing concerns give recruiters something firmer to work with than suspicion.
Security decisions should also involve people with the right expertise. Expecting an HR professional to judge whether an unfamiliar technical warning proves fraud is asking a great deal.
What should HR teams do next?
These stories cover different stages of employment, but they share a common problem: deciding what evidence deserves trust.
For HR teams reviewing their approach to AI, five practical questions are worth asking:
- Are the rules clear? Candidates and employees should understand when AI use is permitted and when disclosure is expected.
- Does the assessment measure ability? A polished application or frequent AI use should not substitute for evidence of good work.
- Can recommendations be challenged? Human review needs enough information and authority to change a decision.
- Is training accessible? Expectations about AI skills should come with reasonable opportunities to develop them.
- Are checks proportionate? Verification should address a defined risk and include a way to correct errors.
For employees and jobseekers, practical AI skills include being able to explain the work produced with a tool’s help. A convincing answer has limited value if you cannot answer the next question.
At AI Tuition Hub, part of our purpose is to help people understand how AI is being used across professions. These October stories show why that knowledge matters in HR: the tools may work quickly, but decisions about recruitment, pay and careers still require informed judgement.
Published: 20th October 2026.