Soft Skills Needed for AI (AIGC): What Malaysian Teams Need to Work With Generative AI, and What Matters More After

Short answer: the soft skills needed for AI fall into two groups. To work with generative AI (AIGC), people need problem framing, critical evaluation of AI output, judgement and accountability, data sensitivity, and learning agility. After AIGC is adopted across a company, the skills that grow in value are the ones AI cannot supply: empathy and trust-building, leadership and influence, collaboration across human-and-AI workflows, original thinking, change resilience, and decision-making under ambiguity.

Most writing on soft skills in the age of AI is framed as a contest between humans and machines. We covered that debate in Soft Skills vs AI, and the durable, tool-independent skills in The Future-Proof Soft Skills Every Malaysian Professional Will Need. This guide is more practical: it is for Malaysian HR, L&D and line managers rolling out tools such as ChatGPT, Gemini or Copilot who need to know which behaviours to train, in what order, and for which roles.

Why Soft Skills Matter in the Age of AI

Generative AI makes a first draft almost free. It does not make it any cheaper to know whether that draft is correct, appropriate for the audience, safe to send, or worth sending at all. That gap between producing output and judging output is filled by soft skills. The importance of soft skills in the age of AI comes down to this: AI multiplies whatever the person using it brings. A clear thinker with good judgement becomes much faster. A careless one produces confident mistakes at scale.

Employer research points the same way. The World Economic Forum's Future of Jobs Report places analytical thinking, creative thinking, resilience and flexibility, and leadership and social influence among the skills employers expect to grow in importance, alongside technology skills such as AI. The technical and the human rise together, not one at the expense of the other.

It helps to separate two moments: the point of use, when one person is trying to get good work out of an AI tool, and the point of saturation, when your whole company and your competitors have the same tools. Different soft skills matter at each, so we treat them as two parts.

Part 1: Soft Skills Needed to Work With AIGC

These are the skills that decide whether an employee gets useful, accurate, safe work out of generative AI, or just more text.

1. Problem Framing: Prompting Is Structured Communication

A prompt is a brief. The things that make a good brief to a colleague make a good prompt: the purpose, the audience, the context the reader does not have, the constraints, and what a good result looks like. Staff who write vague emails write vague prompts and then conclude that "AI is not useful for our work". This is why communication skills in the age of AI matter more than before. Clear writing is now an input to production, not just a courtesy.

At work it looks like: stating the goal in one sentence before prompting, naming the audience (a Bahasa Malaysia-speaking operations team is not a regional head office), and iterating deliberately instead of re-rolling until something looks acceptable. Our Prompt Engineering Masterclass treats prompting as this kind of repeatable communication discipline.

2. Critical Evaluation and Verification of AI Output

Generative AI writes fluently whether or not it is right. The core skill is separating fluency from accuracy: checking claims, figures and references, spotting what is missing, and noticing when an answer is generic rather than specific to your situation. Local accuracy is a common blind spot. General-purpose models often get Malaysian specifics wrong, including regulations, HRD Corp rules, local place names and market conditions.

At work it looks like: a "verify before you forward" habit, tracing every number and regulation to a primary source, and asking "what would make this wrong?" before a document goes to a client. This is root-cause and evidence discipline, the same skill set taught in Critical Thinking & Problem Solving.

3. Judgement and Accountability

"The AI said so" is not a defence to a customer, a regulator or a manager. Employees need to own AI-assisted work as fully as work they wrote themselves, and to judge which tasks should not be delegated to AI at all. A disciplinary letter, a performance conversation or a message to a bereaved colleague may be technically easy to draft with AI and still be the wrong use of it.

At work it looks like: a named human owner for every AI-assisted deliverable, clear personal standards for when AI is and is not appropriate, and the confidence to say "I checked this and I stand behind it", or "I could not verify this, so I removed it".

4. Ethical and Data Sensitivity

Every prompt is a data decision. Customer details, employee records, pricing, contracts and unreleased plans can end up in tools the company does not control. In Malaysia, the Personal Data Protection Act applies to personal data whatever tool processes it. There is also a fairness question: AI-assisted screening, scoring or performance summaries can carry bias that a thoughtful reviewer must catch.

At work it looks like: following the company AI usage policy, anonymising before prompting, never pasting confidential material into public tools, and raising a concern when AI output treats people unfairly. Our Generative AI for Workplace Productivity programme builds these guardrails into the hands-on practice rather than leaving them to a policy document.

5. Learning Agility

AI tools change faster than any annual training calendar. The durable skill is the habit of experimenting, sharing what works, and dropping what does not, even when that means being a beginner again in front of junior colleagues.

At work it looks like: a shared team library of prompts and use cases, and managers who model experimentation openly.

Part 2: Soft Skills That Matter More After AIGC Is Adopted

Once generative AI is everywhere, access to the tool stops being an advantage, because your competitors have the same tool. What differentiates a company then is the human layer around it. These are the soft skills of the AI era that rise in value because AI is common.

6. Empathy and Trust-Building

When routine messages, proposals and replies are AI-drafted, customers and colleagues quickly learn to tell generic communication from genuine attention. Trust shifts to the moments AI cannot handle well: a difficult complaint, bad news delivered honestly, a negotiation where the other side needs to feel heard. Empathy becomes a visible differentiator rather than a background trait.

At work it looks like: handling escalations personally and knowing when to pick up the phone instead of sending a polished AI-drafted reply. EQ in Communication trains exactly these difficult-conversation skills.

7. Leadership and Influence

Managers now lead teams where part of the work is done by AI. They have to redesign roles, set quality standards for AI use, address job-security anxiety honestly, and redefine "good performance" when output volume is no longer a fair measure. Influence matters too: AI can produce the analysis, but a person still has to turn it into a case executives act on.

At work it looks like: supervisors who hold open conversations about how AI changes each role, and managers who can turn AI-generated data into a clear recommendation. See Leadership & Supervisory Skills and Strategic Communication & Storytelling for Leadership.

8. Collaboration Across Human and AI Workflows

A workflow that passed between three people now also passes through several AI steps. Handoffs become the risk: who prompted, who reviewed, who approved. Hidden AI use creates unreliable quality and quiet resentment, so teams need to be open about it.

At work it looks like: explicit agreements on which steps use AI and who reviews them, shared standards across departments, and colleagues who openly say "this section was AI-drafted and I checked it". The fastest way to build this is for a team to build something with AI together, which is the design of our AI-Powered Interactive Team Building Conference.

9. Creativity, Taste and Original Thinking

Generative AI tends to produce the most probable answer, which is often the average one. When every competitor can generate a competent average, the value moves to choosing the right option for this customer and generating the non-obvious idea AI would not offer. Taste, a trained sense of what good looks like, becomes a business skill.

At work it looks like: people who use AI to widen their options and then apply structured judgement to narrow them. Creative Problem Solving & Strategic Decision Making teaches that divergent-then-convergent toolkit.

10. Change Resilience

AI adoption is not one change but a series of them: new tools, new processes, new expectations, often before the last change has settled. Resilience is the capacity to stay effective, and stay decent to colleagues, through repeated disruption without burning out or disengaging. We cover how to train this in adaptability and change readiness training.

At work it looks like: people who ask useful questions about a new system instead of resisting it quietly, and teams that recover quickly after a rollout goes badly.

11. Decision-Making Under Ambiguity

AI gives you more options, more analysis and more scenarios, faster. It does not give you the decision. More options mean more decisions, often made on incomplete information and with trade-offs no model can weigh for you, such as loyalty to a long-standing supplier or the morale cost of a restructure. The skill is making the call, explaining the reasoning, and adjusting when new facts arrive.

At work it looks like: decision criteria agreed before the AI analysis is run, a clear owner for the final call, and the discipline to act on "good enough" information rather than asking the AI for one more report.

AIGC Soft Skills List: Quick Reference

Three Pressures Specific to Malaysian Workplaces

• Multilingual output. Teams prompt and publish across English, Bahasa Malaysia and Chinese. AI translation often looks right and still reads wrong for the audience. Checking register and cultural fit is a communication skill, not a tool setting.
• Hierarchy and face. Junior staff are often reluctant to challenge a superior's work, including an AI-assisted report with an error in it. Critical evaluation only works where psychological safety allows it, which makes it a leadership issue too.
• Personal data obligations. PDPA duties do not pause because a tool is convenient, so data sensitivity must become everyday behaviour in HR, sales and customer service.

How Malaysian HR and L&D Can Build Soft Skills for AI

AIGC soft skills training works best when it is designed alongside the AI rollout, not bolted on afterwards. A practical sequence:

1. Pair every AI rollout with soft skills. Tool training on its own produces faster output without better judgement. Run AI training Malaysia programmes and soft skills training Malaysia programmes as one plan with one set of outcomes. Our breakdown of AI, team building and soft skills explains why the three compound.
2. Sequence Part 1 before Part 2. Start with tool fluency, guardrails and verification (the Part 1 skills), so people can use AI safely. Then invest in the Part 2 skills for the roles where they decide results.
3. Train by role, not by headcount. Customer-facing staff need empathy and verification first. Knowledge workers need problem framing and critical evaluation. Supervisors and managers need leadership, collaboration design and decision-making.
4. Practise on real work, then set norms. Use participants' own documents and customer scenarios, then agree an AI usage policy and review standards so the skills outlast the workshop.
5. Fund it through HRD Corp. Soft skills and AI programmes can be structured as HRD Corp claimable training under SBL-Khas: certified trainer, documented learning outcomes, minimum contact hours, attendance records, and a completion report formatted for your claim. The grant must be approved at least 14 calendar days before the training date under Employer's Circular No. 2/2026, so book three to four weeks ahead. See what makes training HRD Corp claimable.

For why AI literacy itself has become a baseline workforce skill, see Why AI Skills Are No Longer Optional for the Malaysian Workforce.

Frequently Asked Questions

What soft skills are needed for AI? Two groups: skills to work with generative AI (problem framing, critical evaluation, judgement, data sensitivity, learning agility), and skills that matter more after adoption (empathy, leadership, human-plus-AI collaboration, original thinking, resilience, decisions under ambiguity).

Why do soft skills matter in the age of AI? AI makes producing output cheap but not judging it. Soft skills decide whether output is correct and safe to use, and they differentiate once every competitor has the same tools.

What is on an AIGC soft skills list? The eleven skills in the table above, from problem framing to decisions under ambiguity.

Are communication skills still important in the age of AI? More important. A prompt is a brief, and genuine human communication stands out more once routine messages are AI-drafted.

Which soft skills matter more after a company adopts AIGC? Empathy, leadership, human-plus-AI collaboration, original thinking, change resilience, and decisions under ambiguity.

Should we train AI skills or soft skills first? Run them as one plan: AI fluency, guardrails and verification first, then leadership, empathy and decision skills for key roles.

Is AIGC soft skills training HRD Corp claimable? Yes, when structured as training under SBL-Khas with a certified trainer, documented learning outcomes, minimum contact hours, attendance records and a completion report, and with the grant approved at least 14 calendar days before training.

Build Your Team's AI-Era Soft Skills

Redefine Learning Asia is an HRD Corp registered training provider based in Petaling Jaya with more than 25 years of experience, delivering programmes nationwide. We design combined AI and soft skills plans around your roles and real workflows. Talk to us about a programme for your team. 中文读者请看：AIGC时代需要的软技能（中文版）。