AI & Future of Work – JULY 2026

17 July 2026

Future of work

Source: https://www.pwc.com/gx/en/news-room/press-releases/2026/pwc-2026-ai-jobs-barometer.html

PwC 2026 Global AI Jobs Barometer (July 2026)

Sample: PwC looked at over one billion job advertisements posted across six continents. They wanted to understand what is actually happening to jobs and pay in companies that are using AI not what people predict, but what the data already shows.

The most important finding:

AI is not killing jobs uniformly. It is splitting the job market into two different tracks.

Track 1: Jobs that AI is making more valuable: "Professionalised" jobs

Some jobs are being upgraded by AI. AI takes over the routine parts, and the human is left doing the higher-value work — the thinking, the judgment, the leadership.

These jobs are:

Growing twice as fast as other jobs
Paying 42% more than they did in 2021
Increasingly requiring senior-level skills even at junior level

In other terms: a junior employee in a highly AI-exposed role is now 7 times more likely to be expected to demonstrate skills like leadership and strategic thinking than a junior in a low-AI role. The career ladder is compressing. Entry-level is no longer entry-level.

Track 2: Jobs that AI is making easier : "Democratised" jobs

Other jobs are being simplified by AI. AI tools make it possible for almost anyone to do work that previously required training or expertise.

These jobs still exist, but they are growing more slowly and paying less. They are becoming commoditised.

What about the companies using AI the most?

This is where the data challenges conventional anxiety about AI. The companies most exposed to AI are not cutting headcount. They are:

Growing their workforce faster than companies using little AI
Paying higher wages
Achieving 40% higher productivity growth

The top 20% of these companies — the ones who have gone furthest with AI are seeing 163% productivity growth on average. And they are sharing those gains with their employees through higher pay, not just extracting them as profit.

What does this mean?

Companies that use AI well are hiring more people, paying them more, and becoming more productive all at the same time.

The fear that "AI comes in, people go out" is not what the data shows — at least not for companies using AI strategically.

What should HR leaders and CEOs in Malaysia take away?

a) The risk is not job loss, it is skills mismatch.
Jobs are not disappearing. But the skills required to keep them, and to progress in them, are changing faster than most training programmes can keep up with. If your people are not developing judgment, critical thinking, and leadership capability alongside AI tools, they will fall behind even if they keep their job title.

b) Junior talent now needs senior skills sooner.
The traditional "learn the basics for two years then progress" model is under pressure. AI is collapsing that timeline. If you are onboarding graduates or junior hires in Malaysia, you need to rethink what you expect from them — and what you develop in them — from day one.

c) Investing in AI capability is a wage & retention strategy, not a productivity play.
The data shows that companies investing in AI are paying more. If your organisation is not building AI competency, you are at risk of losing your best people to those that are not just in terms of productivity, but in terms of compensation.

d) The window to get ahead is narrow.
PwC found that the gap between high-AI and low-AI companies has tripled since 2022. The organisations that moved early are now pulling away.

New idea : “Orchestration Economy”.

The real winners with AI aren't the tools themselves, but the people who know how to direct them well. Think of a conductor leading an orchestra. The instruments matter, but without someone guiding them, there's no music.

For HR, this changes what "AI skill" means. It's less about knowing every AI feature, and more about judgment: knowing when to use AI, when to check its work, and when to step back and decide yourself. HR pros who can coordinate AI, not just operate it, will be the most valuable ones in this new world of work.

Prompting better

a) Give context, not just a task
Don't just say "write an email." Say who it's for, why it matters, and what tone you want. AI can't read your mind, it can only work with what you give it.

b) Be specific about the output
Length, format, audience, level of detail. "Summarize this" gets a vague answer. "Summarize this in 3 bullet points for a non-technical manager" gets a useful one.

Malaysia AI governance Bill

Source:Malaysia's AI Governance Bill Expands Regulation and Accountability for Businesses | Welcome.AI

Malaysia is drafting its first horizontal legal framework dedicated specifically to AI governance, rather than relying on sector-specific rules. Key points:

Timeline: A Public Consultation Paper was released July 10, 2026 by the National AI Office (NAIO), with stakeholder feedback due by July 31, 2026. Nationwide stakeholder engagement began ahead of a broader public consultation, and the government hopes to table it in Parliament later this year.
Approach: A risk-based approach recognizing AI risks arise throughout a system's lifecycle, with governance responsibilities allocated according to roles and level of control exercised by different parties. Mechanisms include incident reporting, safeguards, and regulatory sandboxes.
Who's regulated: The Bill distinguishes between Developers, who shape the AI system's capabilities, and Deployers, who operate systems in real-world environments.
Five core principles: Human Dignity, Transparency and Explainability, Accountability, Safety and Security, and Data Governance.
New institution: A Central AI Authority would serve as the institutional anchor, overseeing compliance with national standards and cross-sector cooperation.
Content note: Digital Minister Gobind Singh Deo confirmed the bill would not regulate AI-generated content directly, existing laws continue to govern illegal content.
Distinctive IP angle: the bill treats both training-data inputs and AI outputs as intellectual property a structure that sets Malaysia apart from the EU's risk-tier-first model.

Overlapping existing laws already in play

Even pre-Bill, several laws already touch AI use: the Cybersecurity Act 2024 imposes strict incident reporting for national critical infrastructure entities, while the Personal Data Protection (Amendment) Act 2024 introduces mandatory breach notification obligations, and sector regulators including Bank Negara Malaysia, the Malaysian Communications and Multimedia Commission, and the Securities Commission Malaysia each maintain overlapping technology risk expectations creating a regulatory-overlap challenge for compliance teams.

The Bill doesn't yet contain Malaysia-specific employment provisions comparable to the EU AI Act's high-risk classification for HR uses (recruitment, evaluation, etc.) that specificity may come in implementing regulations. For now, PDPA 2010 remains the operative compliance anchor for HR-related AI data use, alongside the general accountability/transparency principles above.