Is now the time companies should commit to responsible AI usage?

Artificial intelligence is no longer a distant technology story. It is already sitting inside search engines, office tools, design platforms, customer service systems, code editors and development workflows. For some people, it feels like a useful assistant. For others, it feels like the beginning of a much larger economic shift: one where human opportunity is quietly traded for short-term efficiency.

The question is not whether AI should exist. It already does. The better question is whether companies should now commit to using it responsibly, before automation becomes the default answer to every business problem.

That is not irrational panic. It is a genuine, answerable question.

We have seen similar patterns before

When automated production lines, industrial robots and computer-controlled machinery entered manufacturing, they did not simply “make work easier”. In many cases, they changed what human labour was needed for. Repetitive manual work, once performed by skilled factory workers, could be carried out faster and more consistently by machines.

The evidence around automation is mixed, but it is not imaginary. Research by Daron Acemoglu and Pascual Restrepo found that industrial robots reduced employment and wages in exposed US labour markets, with one additional robot per thousand workers associated with a lower employment-to-population ratio and reduced wages. MIT’s summary of the research notes that the impact was particularly visible in industries such as automotive manufacturing.

That does not mean every technological advance destroys more work than it creates. Historically, new technology has often created new industries, new roles and higher productivity. The uncomfortable truth is that the benefits are rarely shared evenly. Certain workers move up. Others are retrained. Many are displaced. Some communities never fully recover.

AI may follow a similar pattern, but with one major difference: it is not aimed only at repetitive physical labour. It is now moving into cognitive work — writing, design, coding, legal research, support, analysis, planning and administration.

The employment forecasts are serious enough to deserve attention

The World Economic Forum’s Future of Jobs Report 2025 forecasts major labour-market disruption by 2030: 170 million roles created, 92 million displaced and a net increase of 78 million jobs. That sounds positive overall, but “net increase” can hide a painful transition for the people whose current roles disappear.

The IMF has also warned that almost 40% of global employment is exposed to AI, rising to around 60% in advanced economies. In some cases, AI may complement workers and increase productivity. In others, it may perform tasks currently done by people, reducing demand for labour, lowering wages or removing some jobs entirely.

Goldman Sachs Research has estimated that the equivalent of 300 million full-time jobs globally are exposed to automation by generative AI. Again, “exposed” does not mean every one of those jobs disappears. It means the work can be affected, reshaped, reduced or partially automated.

The responsible position is not to claim that AI will remove every job. It is to accept that the scale of exposure is large enough that companies, governments and workers should be taking it seriously now.

The junior developer problem

Software development is one of the clearest examples. AI coding assistants such as Claude, Codex-style tools, Cursor, GitHub Copilot and other agentic systems can now generate, refactor, explain and test code. In the right hands, they can be very useful, but most developers are aware that they can produce confident rubbish at speed.

The concern is what happens to the learning pipeline.

A senior developer does not appear from nowhere. Senior developers are usually built through years of small mistakes, bug fixes, code reviews, awkward legacy systems, unclear briefs, production issues, mentorship, debugging and gradual responsibility. If companies decide that junior developers are no longer worth hiring because AI can produce “good enough” first drafts, where do future senior developers come from?

Stack Overflow’s 2025 Developer Survey found that 84% of respondents were using or planning to use AI tools in their development process, with 51% of professional developers using them daily. That shows how quickly AI has become normal inside development workflows.

But normal does not automatically mean mature. A 2025 METR study of experienced open-source developers found that, in its trial, developers using early-2025 AI tools took 19% longer to complete tasks than those not using them. The study does not prove that AI always slows developers down, but it does challenge the easy assumption that AI coding tools always create immediate productivity gains.

That is important for responsible AI usage. If AI tools are treated as a replacement for training, mentorship and good engineering judgement, companies may end up weakening the very skill base they rely on.

The real risk is not just unemployment — it is loss of purpose

A society built around work cannot casually remove work without consequences.

Jobs are not only income. They provide routine, status, independence, skill, identity and progression. They give people something to aim for. If stable work becomes a luxury, the question is not simply “how will people pay bills?” It is also “what will people build their lives around?”

Some experts argue that AI will create new work and raise productivity. Others warn that the transition could increase inequality, particularly if the gains go mainly to companies and shareholders while workers carry the disruption. The IMF has specifically warned that AI could worsen inequality if policy does not guide the transition properly.

This is where the debate often becomes too shallow. It is not enough to say, “Technology always creates new jobs.” Sometimes it does. Sometimes it creates different jobs in different places, requiring different skills, while the displaced workers are told to retrain after the damage has already been done.

AI has obvious public-good potential

There are areas where broad public support for AI is much easier to understand. Medical research is one of them. AI is already being explored for drug discovery, diagnostics, clinical trials, imaging, personalised treatment and healthcare administration. The World Health Organization recognises the promise of AI in health, while also warning that it brings governance, bias, privacy and accountability risks.

This is where responsible AI should be easiest to defend: using powerful systems to accelerate medical research, improve diagnosis, reduce administrative burden and support clinicians, while keeping accountability with properly trained humans.

Are world leaders already thinking about this?

Yes — but mostly through the lens of safety, rights and regulation rather than an explicit promise to preserve human career paths.

The UK-hosted AI Safety Summit produced the Bletchley Declaration in 2023, signed by countries including the UK, US, China and EU member states, recognising the need to identify and manage AI safety risks internationally.

The EU AI Act takes a risk-based approach to AI regulation, with rules for developers and deployers of AI systems, including high-risk use cases. It is aimed at safety, fundamental rights and human-centric AI.

The OECD AI Principles promote trustworthy AI that respects human rights and democratic values, while the G7 Hiroshima Process introduced voluntary guidance for organisations developing advanced AI systems.

These are important steps, but there is still a gap. Current frameworks talk a lot about safety, transparency, fairness and risk. They talk less directly about preserving entry-level opportunities, protecting career ladders, and preventing companies from replacing human development with automated output.

That is where a responsible AI pledge could help.

What a responsible AI pledge could include

Companies do not need to reject AI to use it responsibly. A practical pledge could commit them to:

1. Use AI to support workers before replacing them
   AI should first be used to remove repetitive, low-value tasks, not as an excuse to remove people from the business.
2. Protect entry-level roles and training routes
   If junior work is automated, companies should create new supervised pathways where people can still learn, make mistakes safely and develop judgement.
3. Keep humans accountable for important decisions
   AI can assist with analysis, drafting and recommendations, but a responsible person should remain accountable for decisions that affect customers, staff, finances, health, safety or rights.
4. Be transparent with staff
   Workers should know when AI is being introduced, what it is being used for, and whether it may affect their role.
5. Measure real productivity, not hype
   Companies should test whether AI genuinely improves quality, speed and outcomes, rather than assuming it does because the demo looked impressive.
6. Share productivity gains fairly
   If AI increases output, the benefits should not flow only upwards. Reduced hours, better training, improved pay, profit sharing or investment in staff development should be part of the conversation.
7. Avoid replacing judgement with generated content
   AI can produce plausible answers quickly. That does not mean the answers are correct, ethical, secure or suitable.

The question companies should ask now

The responsible question is not:

“How many people can we replace with AI?”

It is:

“How can we use AI to improve the work without destroying the human opportunity behind it?”

Used responsibly, AI could help with medical research, accessibility, safer systems, better services, faster learning and reduced administrative burden. Used irresponsibly, it could hollow out career paths, concentrate wealth, lower wages, remove entry-level opportunities and leave people competing against systems trained on human work.

Companies should not wait until the damage is obvious before taking a position. The time to commit to responsible AI usage is now, while there is still a choice about how this technology is introduced.

AI should be a tool that helps people do better work.

It should not become the reason people are no longer given the chance to work at all.