𝐖𝐇𝐄𝐍 𝐓𝐇𝐄 𝐉𝐎𝐁 𝐃𝐄𝐒𝐂𝐑𝐈𝐏𝐓𝐈𝐎𝐍 𝐒𝐀𝐘𝐒 𝐀𝐈, 𝐁𝐔𝐓 𝐍𝐎𝐓 𝐈𝐍 𝐓𝐇𝐄 𝐉𝐎𝐁

In addition to marking technology jobs, AI will also affect the roles of sales, marketing, legal, HR, education, and operations. The rapid adaptation process is probably the best example. However, the wording of those descriptions says otherwise. This means rejecting a resume excerpt that piqued my attention and letting applicants define what AI literacy, fluency or experience should look like on their own.

Being a marketing manager hired by AI will mean having to do little maths – developing and distributing content. AI can assist with research, but a lawyer must still be accountable for the output, which may be incorrect and is not necessarily confidential or privileged. The processes will partially be automated, but the superhost will control all the generated flows.

Everybody uses AI, but their engagement, expertise, and hazards are different.

This is evident after reviewing the research. According to the researchers, 74% of the ads linked to AI use "AI" in a generic manner, while only 2% referred to ChatGPT. Nearly a quarter of job descriptions contain minimal information on the application of AI. AI had started entering recruitment conversations even before most organisations understood how it would make its way to work.

That means that there should be a job description that explicitly requires AI proficiency. Hence, in order to plan AI, an organisation should be aware of immovable processes, human intervention necessities, data that should be protected and validated, and the impact that AI would make on working processes. Otherwise, experienced candidates won't know that they are competent without any additional notes. The rest will parrot the list of tools used by everyone and provide no evidence of their appropriate use.

This confusion also influences recruitment. TThat certificate is a badge of learning; being familiar with ChatGPT, Copilot, Gemini, or Claude is enough. But that is not going to tell whether an applicant will be able to generate commercial work. The candidate who can explain what problem has been solved, why he used a particular approach, how it has been analysed, and where he disagrees with the machine or vice versa will prove his competence.

The difference is essential as the adoption of AI catches up with the expertise in it. Moreover, 71% of the respondents of a new survey posted by Microsoft and LinkedIn stated that they will be more likely to hire the candidate who lacks experience but has AI skills. This tells about the shift in expectations set – but not about the risky side of it. This polished, AI-created response can very much resemble the expertise of someignorant of how to refute the shaky basis of evidence, lacking the necessary contextual background because of deliberate the bias and skew.

There is always the danger of devolving the interview into a technical test or exam. One way to call to the difference is to provide a report or recommendation generated by AI and a hypothetical customer response, including credible weaknesses. The way he assesses, challenges, protects, and changes the document will tell more than a list of tools on his CV. What is most important is it will show whether I can say, without risking too much, that he understands that the final decision rests with the human.

It is not capable of taking responsibility in the organisation. The best new hire will be able to understand the employer's data, systems, customers, regulatory requirements, and risk appetite. Personal experience with just one tool will never provide you with the necessary organisational context. This is not the best way to go – at least in Europe. The EU AI Act says in Article 4(a) that the provider and deployer of an AI system shall support the development of literacy necessary for persons interacting with an AI system, taking into account, where relevant, their level of knowledge, expertise and experience as well as the context in which they are deployed.

Hence, onboarding is not just about providing the reader with access and policy. It should help him/her to evaluate when AI is useful in their role and when it is not needed. The first commands of the trustful equipment—confidentiality, verification, and accountability—will turn into guided work where judgement has to be evaluated. After 90 days, the new team member should understand process improvement and identify what can be addressed with AI versus what presents risk.

Some European employers have already started treating AI as an organisational competency and not a personal choice. The Lloyds Banking Group has opened the AI Academy to all its 67,000 employees, with training available for everybody irrespective of their position or level of competency. It is not in how many people you train. It understands that employees must integrate AI skills with the context in which they will use them.

All this work is relying on a higher level of integration with performance management. You measure only activity but not contribution – the number of prompts used, platform logins or courses completed, for example. We do not know if AI use improved outcomes, decreased customer value, or reduced errors before decision points, allowing more time for routine and complex tasks. Meta took a step towards solving this problem by adding "AI-driven impact" to the list of core performance expectations starting in 2026.

However, it treats the AI and rewards system in such a way that using this AI without assessing Quality, Verification, and Responsibility will result in unnecessary automation or false output (attributing to so-called "rewards").

Recruitment, hiring, onboarding and performance management

They are all entirely different. But it also tells what extent they are interconnected – thanks in part to AI. Not knowing what the requirements are leads to a bad hire; bad judgement results in the onboarding gap, and this gap appears in the work that needs assessment.

That means that for now, we are not in search of more claimers. The The responsibility is to define responsible AI-enabled work, and there needs to be a cohesive way to bring that together.y are aware the AI will impact work, but they are also explicit about organisations that use those tools to interpret the higher-quality roles.

1. Servet Yanatma, “The Majority of AI Jobs Now Sit Outside Technological Occupations in Europe,” Euronews, July 24, 2026, https://www.euronews.com/business/2026/07/24/the-majority-of-ai-jobs-now-sit-outside-technological-occupations-in-europe.

2. Cory Stahle, “How Employers Are Talking About AI in Job Postings,” Indeed Hiring Lab, October 28, 2025, https://www.hiringlab.org/2025/10/28/how-employers-are-talking-about-ai-in-job-postings/.

3. Microsoft and LinkedIn, “AI at Work Is Here. Now Comes the Hard Part,” Work Trend Index Annual Report, May 8, 2024, https://www.microsoft.com/en-us/worklab/work-trend-index/ai-at-work-is-here-now-comes-the-hard-part.

4. European Commission, “AI Literacy—Questions and Answers,” accessed July 29, 2026, https://digital-strategy.ec.europa.eu/en/faqs/ai-literacy-questions-answers.

5. Lloyds Banking Group, “Lloyds Banking Group Launches AI Academy for 100% AI Literacy by 2026,” press release, January 20, 2026, https://www.lloydsbankinggroup.com/media/press-releases/2026/lloyds/lloyds-banking-group-launches-ai-academy-for-100-percent-ai-lite.html.

6. Jyoti Mann, “Meta Is About to Start Grading Workers on Their AI Skills,” Business Insider, November 14, 2025, https://www.businessinsider.com/meta-ai-employee-performance-review-overhaul-2025-11.