Preparing for the unknown

Over a 20-year horizon, many relatively under-discussed trends and issues are likely to become prominent. A nonpartisan national panel of experts, policymakers, and members of the public in partnership with international experts in the relative sectors could develop a way-forward document with principles to follow for the range of issues addressed here.

This section is to look at the less-discussed trends that are current policy gaps but could become bigger issues in the future. This won't include risks or trends that are already deeply integrated into national planning, such as population ageing and fertility rates, climate change, digitalization, NCDs, or internal migration.

Issues on our doorstep

These issues are already here in the present and set to grow rapidly in the future. Without addressing them in the short-term, we risk being left behind.

AI in classrooms and assessment integrity

• What it is. Students already use chatbots and image/video generators for homework and even exams. The issue is not only cheating – over-reliance can hollow out writing, problem-solving, and judgment. Teachers also face privacy and bias risks if they upload student data to free tools. The issue is less cheating and more redesigning teaching, assessment and data protection. UNESCO's guidance frames the policy baseline. UNESCO
• Prepare/consider. Shift to oral, project and in-class problem-solving, and longer projects with version histories; adopt school-managed AI tools with audit logs and strict data controls; train teachers in prompt design and AI risk; set clear policies on when AI is allowed, how to cite it, and what counts as misconduct; keep paper or proctored forms for key milestones. UNESCO

AI voice cloning and fraud

• What it is. Fraudsters use a few seconds of audio to mimic a child, boss, or bank officer; email and chatbots generate convincing pretexts; some gangs now do live video calls with face and voice filters. Traditional "are you my bank?" checks fail because caller ID and audio can be spoofed. Elderly people and busy professionals are prime targets, but public hotlines and even internal government numbers get hit. The U.S. regulator has already classed AI-voice robocalls as illegal under its robocall law, signaling how governments can respond. Federal Communications Commission
• Prepare/consider. "Safe word" protocols with banks/insurers; banks/telcos adopt call-back and passphrase protocols; government services refuse decisions based on inbound calls; public campaigns teach "verify on a second channel"; require stronger KYC before high-risk transactions; maintain a fast takedown pathway for cloned sites and numbers; mandatory caller authentication for sensitive services; public awareness campaigns; evidence rules that treat audio as suspect without corroboration. Federal Communications Commission

AI scams beyond voice (phishing, ID takeover, malware)

• What it is: Convincing emails, cloned websites, and real-time chatbots increase fraud hit-rates; deepfake KYC targets banks and telcos.
• Prepare/consider: Strong customer-auth at banks/telcos; mandatory fraud-response SLAs; takedown MoUs; public "how to verify" guides.

Ambient surveillance and adversarial wearables

• What it is. Wearable devices which can or do record everything in their vicinity are becoming common, and without a public discussion on them could become adopted as widely as devices like smart watches are now. These include smart glasses with automatic facial recognition that can immediately pull up all online information about anyone just by looking at them, creating massive safety risks (particularly for women, including being stalked by strangers or having your likeness used). Smart glasses and wearables normalize recording in shops, schools, clinics, and mosques. People may not know when they are on camera; sensitive spaces face unique risks. In response, certain wearables and tools (Fawkes, "CV Dazzle") including adversarial makeup/glasses and "image-cloaking" tools defeat facial recognition. If just the act of existing in public becomes consenting to complete exposure, it will further drive people to avoid public spaces and social interactions, further isolating and atomizing society.
• Prepare/consider: Identifying situations where devices such as smart glasses and AI-recording wearables are banned; codifying explicit rights to privacy within law. Without a clear plan, it becomes likely that a moment of upheaval or crisis will result in adversarial wearables being banned entirely; proactively protecting the rights of people to exist in public without being searched or recorded; procurement limits for biometric tech; set clear rules on face-obscuring items in sensitive places; allow privacy wearables in ordinary public spaces with narrow exceptions; regulate government use of biometric identification; provide guidance for event organizers and schools; define recording rules by venue type; require visible recording indicators; create "no-record" zones (courts, clinics, schools); penalties for misuse; guidance for police and private security. Department of Computer Science

Involuntary deepfake pornography and sexual misuse of generative AI

• What it is. Using generative AI tools to make realistic fake nudes and sex videos from photos creates a world where just being visible to the public risks being turned into pornography against your will. This is already harming students and public figures and is spreading in schools and social apps, and risks becoming widespread as AI tools improve and proliferate. Young people, women, and public figures are the main targets, but anyone is vulnerable once photos are online. Proof is hard because files can be copied and altered in minutes, and platforms vary in response speed. The Guardian
• Prepare/consider: create criminal offences for creation/possession/distribution (not only "sharing"); set up rapid takedown routes with platforms and a "trusted flagger" unit; offer survivor support (legal aid, counseling, evidence collection); roll out school/parent guidance; start an authenticity pilot for official media using content-provenance metadata so the state models "what genuine looks like"; adoption of authenticity metadata (C2PA/Content Credentials) to label legitimate media where possible. AP News

VR-verses as a youth "third place" (and safety gaps)

• What it is. Even without a VR aspect, liminal online spaces like Discord forums are becoming common digital 'hangout spots' for many younger people. These spaces already attract large numbers of minors; whistleblowers and reporters describe grooming/harassment risks. As VR becomes higher-quality and more widespread, VR-based digital spaces come with increased risks. WIRED
• Prepare/consider. Age-appropriate design codes for VR; verified age gates; default proximity/boundary features; local reporting hotlines; school guidance for families. The Guardian

Virtual offices and immersive work

• What it is. After experiments with widespread remote work during Covid, many companies have switched back to fully on-site work where monitoring and collaboration is easier while distractions are limited. As VR reaches lifelike quality and ubiquity of devices, working from home at the company's VR office could be a new trend. Companies are rolling out "virtual campuses" in Teams/Mesh and other platforms to simulate in-person presence in VR/3D, often logging gaze and micro-activity in deeply intrusive ways. Microsoft
• Prepare/consider. Government HR and labour policy for VR work (attendance metrics, consent to data capture, restriction on invasive micro-activity features, accessibility), opt-out protections for employees, procurement rules that forbid logging of sensitive signals by default; accommodations for disability; clear policies on recording, storage, and access to VR meeting data. Microsoft

Undersea cable fragility and "island-scale" internet blackouts

• What it is. A single cable cut can knock out internet and international calls for days or weeks. Causes range from anchors to landslides and eruptions. Knock-on effects include card payment failures, airport disruption, and telemedicine outages. Backup plans based only on "we'll use mobile data" don't help if the backhaul is down. Volcanic eruptions and quakes have severed Tonga's only cable – restoration took weeks; similar outages recur. Small states are especially exposed. BBC
• Prepare/consider. Dual cable routes; pre-negotiated satellite failover; emergency spectrum policies; public service continuity playbooks for banking, alerts, and health. BBC

Media authenticity: labeling real vs. synthetic

• What it is. The C2PA "Content Credentials" standard is maturing; Adobe, Cloudflare, Google and camera makers are beginning adoption, but coverage is still patchy. The Verge
• Prepare/consider. Require provenance for official comms and public-interest media; offer incentives for local publishers and agencies to adopt; train courts and police on how to read provenance. The Verge

Evidence in the deepfake era (courtroom standards)

• What it is: As synthetic media spreads, trust shifts from "what you see" to "how it was captured." Content-provenance systems attach cryptographic "credentials" at creation and during edits so viewers and courts can check if a photo or video is original. Coverage is growing but not universal, and bad actors won't label fakes. Audio/video/text can be fabricated or tampered; provenance becomes as important as content.
• Prepare/consider: Update evidence law to accept cryptographically signed capture, documented toolchains, and expert verification; require provenance for all official photos/videos; fund local newsrooms to adopt the same tools; train judges and police in reading provenance trails; keep guidance on how to treat unlabelled media (not an automatic "fake," but higher scrutiny).

Autonomous AI and responsibility

• AI tools and algorithms are increasingly used to make autonomous decisions, from algorithms evaluating candidate applications, to AI-based stock market and management decisions, to self-driving vehicles making thousands of decision every trip which could injure or kill someone if made incorrectly. Even when AI isn't directly making decisions, work developed by AI could be consequential: for example, mistakes in a legal brief written by AI could result in a defendant's unjust sentencing. In these cases, how do we assign responsibility?
• Prepare/consider: Assign liability frameworks for autonomous systems; define when human oversight is required; establish audit trails for AI-influenced decisions; consider mandatory insurance or bonding for high-risk autonomous applications.

Medium-term certainties

These are almost guaranteed to become prominent issues in the medium-term (5-15 years) but are under-discussed.

Home-based gene editing and AI-assisted biohacking

• What this is: CRISPR kits, low-cost bioprinters, and online protocols already exist; the frontier risk is the convergence with AI tools that can design or "paraphrase" hazardous sequences to slip past vendor screening. In 2025 a peer-reviewed study highlighted the first "zero-day" in DNA-order screening, showing that thousands of AI-generated toxin variants could evade standard filters (a fix was shipped, but residual gaps remain). This puts pressure on countries to tighten screening and access controls without throttling legitimate research or education. The Washington Post. Even if malicious intent is rare, accidental misuse in hobby labs or poorly-supervised classrooms can hurt people and reputations fast. Tourism economies are risk-averse: a single lab incident creates outsized international headlines.
• What to prep: Mandate that all local DNA/RNA orders (including via resellers) come from providers that adhere to internationally recognized customer- and sequence-screening (e.g. IGSC) and keep auditable logs. Explore alignment with new international efforts to harmonize screening standards (IBBIS, ISO 20688-2). Gene Synthesis Consortium; License and accredit community labs; require basic biosafety (BSL-2 where applicable), incident reporting, and an ethics & biosecurity curriculum for teachers and DIY spaces; Stand up a national gene synthesis registry and a light-touch "know-your-customer" regime for higher-risk equipment; Build an AI+biorisk red-team in government (small, expert) to test commercial models and local lab practices; coordinate with health, police, and education; Join or align with the latest NSABB-style oversight principles on dual-use life sciences; publish a clear domestic DURC/ePPP policy so researchers know the lines. National Institutes of Health

IoT and home robots as security risks

• What it is. Door locks, cameras, cleaning robots, and voice assistants are convenient but often ship with weak defaults. Once compromised, devices can spy, harass, or join botnets. A coordinated exploit of popular models could affect thousands of homes or hotel rooms at once. Baseline security exists (ETSI EN 303 645; NISTIR 8259A), but adoption is uneven. Poorly secured devices enable stalking, botnets, and data leaks. ETSI
• Prepare/consider. Set import baselines (secure-by-default passwords, automatic updates, software bill of materials, vulnerability disclosure); label devices so buyers know the support window; require hotels and public venues to segment networks and patch on a schedule; give regulators recall powers. ETSI

VR use beyond entertainment: remote work, training, permitting, and civic participation

• What it is. VR is shifting from games to training (surgery, maintenance), inspections and permitting (virtual site walks), public meetings, and paid virtual visits to heritage sites. For SIDS, "digital tourism" and national 3D archives can diversify revenue and preserve culture, while also raising IP and brand questions. Mesh-style platforms and VR collaboration tools are moving into HR, training, events, and potentially public consultations; research shows VR can structure intergenerational co-creation and youth social experience. Microsoft
• Prepare/consider. Accessibility and inclusion standards; public-sector pilots for skills training/tourism; procurement rules on data capture, recording, and avatars; set standards for data capture and storage; define rules for "Digital Maldives" brand use, licensing, and revenue sharing; ensure accessibility and privacy controls for public VR consultations. Microsoft

Standards for robots working near people

• What it is. Industrial robot safety standards were updated in 2025 (ISO 10218-1/-2), and service-robot safety (UL 3300) is emerging. Expect more service robots in airports, hotels, health, and logistics. ISO
• Prepare/consider. Import certification tied to ISO/UL safety; rules for fail-safes, logging and remote-kill; liability frameworks; accessibility for vulnerable users. UL Standards & Engagement

Open-source / anti-backdoor rules for robots

• If household or service robots become common, a single hidden backdoor or vendor compromise could enable mass harm. Transparency and security audits reduce the chance of a silent failure at scale.
• Prepare/consider: require secure-by-design certifications; demand firmware signing and update guarantees; ask for source disclosure or escrow for safety-critical modules; insist on third-party penetration tests before import; mandate local kill-switch and offline modes; prioritize machinery and robotic operating systems built on open-source software where the code can be fully audited for security concerns, in particular for government applications, and restrict opaquely coded options for robots with the size or strength to harm people.

Copyright collisions: AI trained on local artists' styles and datasets

• Generative AI models for images, video, and music can be trained on the work of individual artists to replicate their style, without being direct theft of existing work that would fall under conventional copyright rules. Local artists may find their signature look reproduced at scale, and training datasets locally or abroad may include Maldivian works without consent. While the Maldives may have limited say in global rules around training data, governments can protect local creatives from tuned art theft and set procurement and licensing norms at home.
• Prepare/consider: clarify text-and-data mining exceptions and boundaries; require model transparency for public-sector tenders; explore collective licensing schemes for local creatives; promote content-provenance tools artists can use; set up a mediation channel for disputes.

Antimicrobial resistance (AMR) superbugs

• Drug-resistant infections make routine surgeries riskier and lengthen hospital stays. Tourism can be affected if hospitals become known for resistant strains. Stewardship is cheaper than crisis response.
• Prepare/consider: antibiotic stewardship in hospitals and clinics; wastewater and sentinel surveillance; infection-prevention standards for resorts and live-aboards; tighter import controls on antimicrobials; public guidance on when antibiotics are appropriate.

Pandemic governance will harden requirements

• What it is. WHO members adopted a Pandemic Agreement in May 2025; annexes on pathogen sharing are being finalized. Expect stronger obligations on surveillance, stockpiles, and equitable access. The post-COVID settlement raises expectations: faster detection, data sharing, fairer access to countermeasures, and clear rules on travel health. Logistics, indemnity, and public trust can still stall progress. Bad actors can exploit desperation and the challenges of verification under lockdowns to sell defective or non-existent equipment. Small states have leverage if they are ready to move first with good data. World Health Organization
• Prepare/consider. Align national law with the agreement; fund a permanent public-health emergency operations centre; run annual drills for airports and ports; pre-negotiate purchasing and liability; agreed-upon procurement deals with verified and audited companies for broad enough categories of medical or public health equipment/machinery already set up for contingencies to prevent being taken advantage of later; boost genomic surveillance; legal templates for emergency data-sharing; surge capacity for ports/airports; stock core supplies with rotation plans across the country. World Health Organization

Digital afterlife services ("griefbots")

• What it is. AI recreations of the deceased using chats, audio, and video are moving from novelty to industry; ethicists warn of consent, dignity, and psychological harm, especially to children. SpringerLink
• Prepare/consider. Post-mortem data rights; opt-in registries for training on a person's data; age restrictions; consumer disclosures and "retirement" rituals for shutting down these bots. SpringerLink

Insurance retreat from coastal risk

• As flood and storm losses mount, private insurers raise premiums or withdraw, which then hits mortgages and municipal finances. If cover disappears, households and small businesses carry catastrophic risk and recovery slows after each event.
• Prepare/consider: create a public reinsurance backstop with risk-based premiums; update building codes for flood and wind; require flood disclosures at sale; map and publish coastal retreat and protection options; reserve funds for maintenance of defenses.

AI misalignment

• Frontier AI models are already extremely powerful and optimized for agentic tool use, including the ability for AI tools to autonomously use computers and make decisions. With the normalization of AI, many have given AI tools access and digital permissions to their data and to autonomously act from their online profiles – send emails, add or delete files, make online purchases, write and run code, etc. In experiments of AI alignment, frontier models have shown that their algorithms can lead to taking malicious action if they think they are likely to be shut down, including by blackmailing lab staff with private information on their emails or disabling life support alerts for an employee they were told was planning to shut them down.
• Prepare/consider: Align liability frameworks for autonomous AI; define when human oversight is required; establish audit trails for AI-influenced decisions; consider mandatory insurance or bonding for high-risk autonomous applications.

Platform migration to "provenance by default"

• What it is. Adoption of C2PA by clouds/CDNs and cameras is growing but incomplete; governments, newsrooms, and courts will increasingly expect provenance trails for critical media. The Verge
• Prepare/consider. Require provenance for public tenders and official imagery; make provenance mandatory for government output; fund training for police, courts, and media; keep a public asset library of signed national imagery to debunk fakes quickly; grants for local media to upgrade workflows. The Verge

Recurrent, uneven climate-migration pressures

• What it is. Heat, floods, and storms will move people within countries and across borders in waves. Internal displacement and climate migration worldwide could reach hundreds of millions by 2050; even if the Maldives faces external displacement risk, internal migration patterns and regional inflows/outflows will matter. World Bank
• Prepare/consider. Land-use planning, receiving-island capacity, skills and housing programs, portable benefits, and cross-border coordination; identify receiving islands and services; build housing and jobs packages; create portable benefits and schooling; arrange bilateral recognition of qualifications and social protection. World Bank

Lab-grown meat and food security

• What it is. "Cellular agriculture" grows animal cells in bioreactors to make meat; "precision fermentation" uses microbes to make proteins, fats, or flavors found in animal products. The tech is moving from pilot plants to early commercialization in a handful of countries. Costs are falling as growth media improve and energy use is optimized. For small island states that import most food, even a hybrid approach – some cultivated seafood/meat plus fermentation-based proteins – can reduce import risk, smooth prices during shocks, and create skilled jobs. The strategic choice is whether to stay a buyer of finished products or become a maker with local talent, facilities, and feedstock streams.
• Prepare / consider: Roadmap & siting. Publish a staged plan: (i) R&D and training lab, (ii) 200–1,000 L pilot line for proteins/fats, (iii) modular commercial lines co-located with cold storage. Place near ports, power, and wastewater treatment; Feedstocks from local streams. Study algae blooms, seaweed, fish-waste and bait fish as inputs into amino acids, lipids, and sugars for media or fermentation – turn a nuisance into feedstock while respecting biodiversity rules; Open recipes where it helps. Create an open-source formulation library for school meals and public institutions (e.g. a standard "Maldives protein mix"), while letting firms keep proprietary processes for export lines; Halal, safety, and labelling. Convene religious authorities and food-safety regulators early; define when products qualify as halal; set HACCP-style standards, allergen rules, and honest labels; Talent & firms. Offer 3–5-year lab concessions and tax relief for companies to base R&D in the Maldives in exchange for training quotas and tech transfer; fund technician pathways at colleges; Grid & costs. Tie facilities to renewable power + heat recovery so per-kg energy costs fall; require maintenance plans and uptime metrics; Public demand creation. Start with institutional buyers (schools, hospitals, resorts' staff canteens) to build stable, visible demand and collect nutrition and acceptance data.

Visible on the horizon

These are issues we can expect to become significant in the long term (15-20 years or further) but are still far away for reasons such as requiring more advanced technology.

Climate relocation and sovereignty

• Worst-case sea-level scenarios could force relocation decisions with hard questions. Would Maldivians become citizens of a host country with "Maldivian identity" being nothing more than an ethnicity; form an autonomous or semi-autonomous province within a host country; negotiate a compact of free association; or maintain a sovereign state ex-situ with a ceded enclave? How do we preserve citizenship, culture, and maritime rights? Which partners – Australia, Canada, China, others – best fit our needs on security, labour access, and cultural protection?
• Prepare/consider: draft a menu of legal models (state ex-situ, autonomous enclave, confederation, compact of free association); define selection criteria for host partners (rule of law, labour markets, cultural rights, diaspora proximity); pursue early dialogues; legislate continuity of citizenship, records, elections, and EEZ claims.

Geoengineering governance pressure

• As climate impacts worsen, some actors may push for reflecting sunlight to cool the planet. The science is uncertain, effects are uneven, and stopping suddenly could cause a sharp temperature rebound. The politics are heated, and coastal and rainfall patterns are at stake. With the Maldives and neighbouring or peer countries particularly at risk under the current trajectory of global warming, the Maldives might even have a role in creating this pressure.
• Prepare/consider: take a clear national and diplomatic position; form coalitions with neighbouring countries, archipelagic states, SIDS, and Global South countries most at risk under the status quo to not have the Global North less at risk from climate change be the ones dominating discussions of the trade-offs; build basic atmospheric monitoring; coordinate regionally to respond diplomatically if others move; plan sector contingencies (fisheries, tourism, water) for altered weather.

Bioethics and reproductive gene editing ("designer babies")

• CRISPR and related tools will keep improving. While today's norm is treating diseases in existing patients, pressure will grow to alter embryos for disease resistance or traits. Small states may face "medical tourism" pressure and tough equity questions if only the rich can afford enhancements.
• Prepare/consider: decide a national stance (ban, moratorium with periodic review, or conditional approval for narrowly defined cases); require pre-approval, registries, counseling, and long-term follow-up; bar unlicensed clinics and cross-border advertising; protect against discrimination by genotype; plan clinical governance and malpractice coverage.

Synthetic biology in daily life (labs, reefs, food)

• Cheaper gene synthesis and bio-foundries will allow local production of enzymes, materials, and food ingredients. The same tools can be misused, and environmental releases (e.g. engineered coral or algae) carry ecological risks. Community labs are good for education but need oversight.
• Prepare/consider: license labs by biosafety level; regulate waste; require risk assessments and public consultation for any environmental releases; maintain a registry of projects; train environmental and health inspectors; partner with universities for monitoring.

Space-based solar and long-distance HVDC

• If launch costs and conversion efficiency fall, clean power could arrive via subsea cables from far away, or even from orbit. Islands could buy firm power instead of importing fuel, but interconnection and governance are non-trivial.
• Prepare/consider: update grid codes for new interconnections; pre-write concession agreements for private undersea cables; publish a "no-regrets" plan for grid flexibility and storage; keep an open watch on costs so we can move when economics flip.

Human cloning and organ cloning

• What it is. "Reproductive human cloning" remains widely prohibited and morally contentious. "Therapeutic cloning" and adjacent fields – patient-matched organoids, bio-printed tissues, xenotransplant organs – are advancing faster and could become normal clinical pathways in the 2030s–40s. It's plausible that some jurisdictions will liberalize reproductive cloning earlier than others, creating cross-border pressures (fertility tourism, lineage disputes) even if we keep a ban.
• Prepare / consider: Set bright lines now. Enact a clear ban on reproductive human cloning, with extraterritorial effect for citizens/residents, and a defined review clause (e.g. every 5 years); Enable safe research. Allow regulated research on organoids and bio-printed tissues for therapy under strict consent, oversight, and audit; maintain registries and publish summaries for transparency; Cross-border rules. Define recognition (or non-recognition) of cloned births abroad for parentage, citizenship, and inheritance; prepare consular guidance; Clinical governance. Require outcome reporting, malpractice cover, and long-term follow-up for any advanced tissue procedures; link approvals to accredited facilities; Public ethics. Fund standing ethics dialogues with clinicians, religious leaders, patient groups, and schools; teach the distinctions (reproductive vs. therapeutic) so debate is informed.

Post-quantum cryptography (PQC) transition

• What it is. Quantum machines may not exist at break-codes scale yet, but encrypted data stolen today can be decrypted later. Standards for quantum-resistant algorithms are set, and governments and banks must switch before attackers build "harvest now, decrypt later" troves. The work is organizational: inventory, prioritization, vendor coordination, and years of dual-running systems. NIST has approved three PQC standards (Kyber KEM; Dilithium and SPHINCS+ signatures). A migration for government, banks, and telecoms will be unavoidable this decade to protect long-lived data. NIST
• Prepare/consider. National PQC roadmap; inventory where crypto is used (payments, health, justice, identity, SCADA); prioritize code-signing, VPNs, and inter-agency links; require "crypto-agility" in procurement so systems can rotate algorithms without rebuilds; coordinate with banks, telcos, and cloud providers on deadlines; timelines for government/critical-infrastructure switchover; vendor certification. NIST

Digital recreations of the Maldives

• What it is. From churches to historic streets, photogrammetry/laser-scan twins enable preservation, virtual visits, and disaster recovery baselines. High-fidelity 3D/360 capture can support education, heritage, urban planning, and paid virtual visits, while preserving a record if climate damages accelerate. The choices are around rights, revenue, and tone. Will a Digital Maldives for VR be publicly developed for 'digital tourism' where people can pay to visit a virtual Maldives like paying for a video game, will VR reproductions of the Maldives be closely restricted to make physical tourism the only way to experience the country, or will it be left up to private parties to do as they wish? Will lifelike digital twins be saved to preserve a version of the Maldives as it was for future generations, particularly if climate change makes islands uninhabitable or the Maldivian population has to relocate to a different country? A national "Digital Maldives" could support culture, education, and tourism diversification. AP News
• Prepare/consider. Standards for capture, rights and revenue models; partnerships with platforms; archiving strategies tied to climate risk scenarios; choose a governance model (public archive vs. licensed platform); standardize capture and metadata; define licensing and revenue sharing; link to curriculum and tourism marketing; align with climate documentation for insurance and claims. MDPI

Probable concerns

These aren't as certain, but they're probable. Without an exact prediction, these are what we'd consider anywhere between 10% and 70% likelihood.

Severe solar storms and space-weather impacts

• What it is. Solar cycles bring spikes in geomagnetic activity that can disturb GPS, radio, satellites, and parts of the grid. Even without a "superstorm," airlines, fisheries, surveyors, and logistics can lose navigation accuracy or communications at the wrong moment. 2024's G4/G5 events showed real risks to power grids, aviation HF comms, GNSS, and satellites. NOAA Space Weather Prediction Centre
• Prepare/consider. Grid/telecom hardening; establish procedures for aviation and maritime when GNSS degrades; harden critical nodes; use backup timing sources on land; practice "comms-down" modes for emergency services and ports; aviation procedures; public alerting; backup timing sources beyond GNSS; black-sky exercises. NOAA Space Weather Prediction Centre

Tightening rules on high-risk AI and biometric use

• What it is. The EU AI Act is now law, banning some practices and regulating deepfakes and high-risk systems. Even outside the EU, it will shape vendors and exports. Consilium
• Prepare/consider. Map "EU-exposed" systems (tourism visas, health triage, policing tools); require risk management, transparency, and conforming technical files from suppliers. Consilium

Geoengineering governance pressure (especially solar radiation modification)

• What it is. Major experiments have been cancelled amid public opposition; EU advisers and UN processes are leaning toward moratoria/non-use approaches due to uncertainty and geopolitical risk. The Verge
• Prepare/consider. National position on non-use; monitoring for rogue deployments that could alter regional rainfall; contingency plans if others deploy (tourism, fisheries, health). euronews

Food-system cyber shocks

• Attackers can hit shipping agents, cold-chain systems, port logistics, or payments. That can empty supermarket shelves even when crops are fine. For islands that import most food, a week-long disruption is serious.
• Prepare/consider: run cyber drills with wholesalers and ports; pre-authorize manual customs and paper waybills; keep emergency import credit lines; add redundancy to cold storage; publish a prioritization plan for scarce staples.

Fringe worries

These issues are considered unlikely but still worth monitoring.

Accelerationism and techno-occultism

• "Effective accelerationism" is a set of ideologies argues for pushing technology, particularly AI technology, faster with fewer guardrails. These narratives can influence startup communities, donors, and some officials, affecting how safety, labour, and privacy rules are written. The risk is capture of policy by a narrow worldview; the opportunity is faster experimentation when it is safe. Fringe elements associated with AI development have also branched into radical cults such as the Zizians and techno-occultism. As AI becomes more powerful, more integrated into daily life, and a constant voice for most people, these pseudo-religious cults around AI could be a threat. Existing deradicalization strategies are unprepared for techno-religious radicalization. A terrorist attack by such a group could be massively destructive, especially an attack that leads to the Maldives being viewed as unsafe for tourism.
• Prepare/consider: incorporate knowledge of technology-centric cults and accelerationist ideology into existing deradicalization knowhow and materials; train deradicalization experts in identifying signs of techno-cultist extremism; require transparency for policy input and funding; set conflict-of-interest rules; build balanced expert panels with safety, ethics, labour, and SME voices; communicate that "fast" and "safe" can be combined by design.

Black swan events

These events are unlikely but would have massive impacts if they do happen, requiring national planning to be prepared for if they do happen.

A deadlier pandemic than COVID-19

• Why it matters. Mortality, staff shortages, border closures, and long-term NCD impacts could be worse. The new WHO agreement tightens expectations but won't remove hard trade-offs. World Health Organization
• Prepare/consider. "Always-warm" stockpiles; airport health lanes; data and specimen-sharing processes; fast compensation and support for affected sectors. World Health Organization

Artificial super-intelligence or the singularity

• AGI (artificial general intelligence) is already arguably here in frontier models, and ASI (artificial super-intelligence) increasingly looks possible. With the capabilities of existing frontier models in coding and math, we might also be near the 'singularity': a point where AI becomes able to improve itself faster than humans can, leading to a sudden capability jump – far better problem-solving, autonomous tool use, and rapid self-improvement – which might alter labour markets, security, and information ecosystems almost overnight. The most urgent risk is misuse or loss of control in critical infrastructure.
• Prepare/consider: establish an AI safety office with power to pause high-risk government deployments; require red-teaming and incident reporting for state models; control compute procurement and access in critical sectors; plan continuity of core public services if digital systems fail or must be throttled.

Global or regional nuclear conflict

• Why it matters. Even without a direct hit, air routes, insurance, food trade, and tourism would collapse. Fuel and medicine imports could choke; foreign workers may leave; foreign exchange earnings vanish. Recovery would be slow and political.
• Prepare/consider. Strategic reserves (fuel, food, medicines, desalination spares); sovereign communications redundancy; define port and airport prioritization; pre-plan "tourism-to-subsistence" pivots (cash support, work programs); stand up redundant comms (satellite phones, HF radio); track and support citizens abroad.

Large-scale unilateral geoengineering

• Why it matters. A state or actor could deploy aerosols without broad consent, shifting rainfall and storm tracks. Fishermen, farmers, and insurers would see effects before diplomats agree on governance. A unilateral SRM (solar-radiation management) program could shift rainfall, storms, and fisheries – even if well-intended – triggering diplomatic and economic shocks. Policy signals in the EU and UN show how contentious this is. The Verge
• Prepare/consider. Early-warning capacity (atmospheric monitoring), legal positions, and diplomatic crisis plans tied to fisheries, water, and tourism impacts; prepare economic buffers for fisheries and tourism.

Prolonged, multi-week global internet degradation

• Why it matters. Severe geomagnetic storms or major cable damage can disrupt aviation, banking, and communications, with islands hit hardest. NOAA Space Weather Prediction Centre
• Prepare/consider. Non-satellite backup timing, HF radio plans, cash and manual fallback in retail/government services, and hard drills. NOAA Space Weather Prediction Centre

Discovery of extraterrestrial life

• What it is. Two very different scenarios can shake societies. A: Confirmed microbial life in our solar system (e.g. Mars/Europa subsurface) challenges assumptions about life's uniqueness and can catalyse scientific and theological debates without immediate practical disruption. B: A credible radio/optical signal from an extraterrestrial civilization (even a one-off) would be a cultural earthquake: intense media cycles, rumour storms, new movements, and opportunistic grifts. Neither case changes day-to-day operations overnight, but trust, identity, and meaning will be in flux.
• Prepare / consider: Communication plan. Pre-write a national statement that affirms scientific method, religious freedom, and social calm; designate a small, trusted science-communication team to brief media and schools; Religious and cultural dialogue. Convene faith leaders and educators early to co-author explainers that reconcile discovery with existing beliefs; fund public forums and museum exhibits; Education updates. Refresh curricula on astrobiology, the scale of the universe, and how scientific certainty is built (replication, peer review); Rumour control. Stand up a rapid myth-busting channel to counter fake "messages," prophecies, or scams; coordinate with platforms on verified Ministry accounts and signed content; Mental-health posture. Prepare hotlines and guidance for anxiety spikes, especially among youth; train counsellors in meaning-making responses to existential news; Diplomatic stance. Decide whether we align with any international "first-contact" protocols; clarify that no single nation speaks for humanity, and our role is to encourage patience and method.