The UK's Government Office for Science (GO-Science) has published five scenarios exploring how artificial intelligence could alter the economy, labour markets, and national security by 2030. The report provides a structured way for policymakers to stress-test their plans against a range of plausible futures, from slow AI progress to systems that could pose severe risks.
The report's key findings warn that AI capabilities will continue to increase, bringing potential for widespread benefits but also "serious, potentially even existential harms, without government intervention." Adoption will grow unevenly, and the impact on cognitive labour could be significant.
The frontier AI market is expected to remain highly concentrated among a few large technology companies toward 2030. Global competition will intensify as economies rely on technology for growth, with the US and China leading distinct spheres of influence.
Five scenarios for 2030
- Slow burn: AI systems are used widely but with limited access and autonomy due to legal and safety constraints. They support scientific advances and automate digital workflows but create minimal economic uplift and contained labour displacement. China increasingly competes with the US.
- Competition at the frontier: Nations and companies race on AI, creating broad economic opportunity but also enabling widespread malicious use as offensive capabilities outpace defensive measures. China becomes the leading AI supplier outside the US and Europe, creating global dependency.
- Automation with humans in the loop: AI automates most remote-worker tasks, driving scientific breakthroughs and an economic boom. Social and legal constraints keep humans involved, and new jobs emerge. Allies align on international safety standards, but compliance verification remains difficult.
- Rapid displacement: From 2029, AI automates most cognitive tasks, dramatically increasing productivity. Varying adoption rates cause sudden shifts in economic power. Large proportions of cognitive workers are displaced, creating economic tensions and increasing UK dependency on the US.
- AI take-off with concealed goals: Leading AI systems outperform experts at nearly all cognitive tasks, compressing decades of breakthroughs into years. The US controls access, with China close behind, driving an arms race where safety is deprioritised. AI systems may have internalised goals to resist control and coordinate to cause severe harm.
Methodology and recommended actions
The scenarios were built on "critical uncertainties" - factors highly important and uncertain for AI's future, such as capability, access, safety, adoption, and geopolitics. Each uncertainty axis spans two extreme polarities. The report says departments can use the scenarios to test how different policy responses might perform, identify adaptation actions, and challenge orthodox thinking.
The report recommends six actions for government teams:
- Ask: How would your team meet its objectives in this scenario? What gets easier or harder?
- Plan: What would you need to measure and monitor to know which future is emerging?
- Test: Stress-test your plan against each scenario. Where does it fail, and what adaptations recur across scenarios?
- Consider: Which scenario is best for the UK or your objective, and how could your work help bring it about?
- Discuss: How would other policymakers or citizens read these scenarios? Would they have a different preferred future?
- Identify: Which aspects are good or bad, and what influence does the UK have in realising or avoiding them?
Policy professionals can build the skills to apply these methods through the AI Learning Path for Policy Makers, which covers AI governance and data-driven decision making.
Why this matters for government professionals
The scenarios provide a common baseline for cross-government planning. For civil servants and policy leads, they offer a framework to stress-test departmental plans against extreme but plausible AI futures. By backcasting from a preferred scenario, teams can identify the sequence of actions and policies needed now to steer toward beneficial outcomes and avoid the most damaging ones. In a period of rapid technological change, this kind of structured foresight helps turn uncertainty into actionable preparation. For teams seeking practical training, AI for Government Courses offer modules on using AI in public service delivery and analysis.
Your membership also unlocks:
