Are gig workers left without a safety net as mental health claims and artificial intelligence reshape pay cover?

The idea of a financial safety net at work used to be easier to picture: if someone fell ill, lost hours, suffered an injury, or needed time away from work, there was at least a framework of sick pay, insurance, or state-backed support to lean on. For many gig workers, that picture is far less clear. As platform-based work grows and technology plays a bigger role in setting tasks, monitoring performance, and even removing access to work, an important question is becoming harder to ignore: are gig workers being left exposed just as the risks around income and wellbeing are becoming more complex?

Recent evidence suggests the concern is not theoretical. Across multiple international studies, a pattern is emerging of weak social protection, rising mental-health strain linked to insecurity and surveillance, and AI systems that can affect pay, access to work and deactivation with limited human oversight. For UK readers, the details may vary by employment status and provider, but the broader lesson is highly relevant: when work becomes more flexible, protection often becomes more fragile unless people actively review what cover they do and do not have.

The safety net was already thin before AI entered the picture

One of the clearest findings from recent labour research is that many platform workers are still operating with limited access to core protections. The International Labour Organization reported in April 2025 that around 40% of surveyed web-platform workers in Latin America and the Caribbean were excluded from health and social security coverage. Its broader survey report, covering 1,153 workers across 21 countries in the region, described a recurring mix of precariousness, informality and lack of social protection.

That matters because gig work is often presented as flexible and accessible, but flexibility on its own does not replace structured protection. A joint ILO, OECD and ISSA policy paper found that platform workers’ vulnerability was starkly exposed during COVID-19, especially where they lacked access to unemployment insurance, job-retention schemes, sickness benefits and health protection. The same paper said coverage remained shallow across many risks, with only about 20% of platform workers having old-age protection, while cover for work accidents and unemployment was limited.

The issue also extends beyond one region or one crisis. More than 90% of EU digital labour platforms classify workers as self-employed, according to the European Commission. While self-employment can be appropriate in some cases, the Commission said about 5.5 million of an estimated 28 million people working through platforms in the EU in 2021 may have been misclassified. That classification question is central because it often determines access to labour rights, social benefits and practical support when income suddenly stops.

Low and unstable earnings make protection harder to maintain

Income protection, savings and insurance all depend, in one way or another, on income that is regular enough to plan around. Yet recent data shows how fragile earnings in online gig work can be. The ILO’s 2025 regional survey found median hourly income of just USD 2.57 and average hourly income of USD 5.48 among surveyed web-based platform workers in Latin America and the Caribbean. Even allowing for differences between regions and cost of living, the underlying message is clear: many people in platform work are earning too little to build meaningful financial resilience.

There is also the question of consistency. In the same ILO survey release, 52% of respondents said platform work was not their main source of income. On one hand, that suggests gig work is often supplementary rather than a full replacement for conventional employment. On the other hand, it also means many workers may drift into a grey area where neither their main job nor their side income gives them complete protection, especially if they assume flexibility itself offers security.

For households trying to budget, this instability can become a serious practical problem. If income rises and falls month to month, it is harder to commit to protection products, emergency savings or long-term retirement planning. It can also make it easier to delay decisions about cover until a crisis arrives. In practice, the lack of stable pay can turn the absence of a safety net into a cycle: low earnings reduce the ability to pay for protection, and limited protection makes income shocks even harder to absorb.

Mental health is becoming a core financial risk, not a side issue

Mental health at work is now too significant to treat as separate from financial planning. The World Health Organization says 15% of working-age adults were estimated to have a mental disorder in 2019, and that 12 billion working days are lost every year to depression and anxiety, costing about US$1 trillion annually in lost productivity. Those are large global figures, but they point to something very personal: when mental health suffers, income can suffer too.

The WHO’s list of workplace mental-health risks is particularly relevant to gig workers. It includes excessive workloads, low job control, job insecurity and inadequate pay. Those features are often discussed as ordinary parts of platform work, yet they are also the same conditions linked to poorer mental wellbeing. When someone is never fully sure how much they will earn next week, whether a rating will affect access to work, or how long a platform account will remain active, the strain is not only emotional but economic.

The WHO also highlights a wider structural gap: more than half the global workforce works in the informal economy, where there may be little or no regulatory protection for health and safety and limited access to social or financial protections. That framing matters because it shows why mental-health pressures in gig work cannot be viewed only as personal resilience issues. They are often rooted in how work is organised and in whether workers have meaningful support if their health affects their ability to earn.

AI and algorithmic management are reshaping the day-to-day reality of work

Artificial intelligence is no longer a niche issue in employment. According to the European Commission’s Joint Research Centre, 30% of EU workers were using AI on the job in 2024/2025. The same JRC research found that 42% of EU workers fall under what it calls “partial platformisation”, meaning they are exposed to at least one form of digital monitoring and one form of algorithmic management. In other words, the systems once associated mainly with app-based gig work are now spreading more widely across the labour market.

Digital surveillance is already common. The JRC said 37% of EU workers were digitally monitored for working hours and 36% for entry and exit times. For gig workers, monitoring can go further, affecting task allocation, ratings, pricing, incentives and account standing. Recent UN DESA analysis warned that algorithmic management is already contributing to work intensification, increased digital surveillance, biased decision-making and reduced human oversight in employment decisions.

This matters financially because control over work is shifting away from direct human managers and towards systems that can be difficult to question. If an algorithm influences how much work someone sees, what jobs they are offered, or how their performance is judged, then technology is shaping earnings and security in a very immediate way. That is one reason current policy debate is moving beyond wages alone to the wider architecture of benefits, data, automated decision-making and workers’ rights.

AI may not only change work, but also reduce what some workers are paid

There is a further twist in the AI debate that deserves attention: using AI may affect what workers are paid, even when the quality of their output does not change. A 2025 paper covering 13 studies with 4,956 participants found that people consistently lowered compensation for workers who used AI compared with those who did not. The researchers said this “AI penalty” also appeared with real gig workers and real monetary compensation, suggesting it could influence live freelance and platform markets rather than existing only in hypothetical surveys.

If that pattern holds more broadly, it raises difficult questions about pay cover in digitally mediated work. A worker may adopt AI tools to improve speed or efficiency, only to find the market discounts their labour because of how that work is perceived. In simple terms, technology could increase pressure to produce more while also weakening pricing power. For workers who already lack guaranteed pay, employer sick pay or robust benefits, that creates another layer of uncertainty.

For anyone relying on freelance or platform income, this reinforces the importance of not assuming future earnings will remain steady just because demand exists today. AI can shift customer expectations, competition and pricing very quickly. From a financial planning perspective, that means it is sensible to think not only about present income, but about how vulnerable that income could be to technological change and changing market attitudes.

Deactivation can turn a working week into an income shock overnight

One of the most severe weaknesses in gig work is how abruptly income can disappear. A 2025 research paper on platform work described deactivation as the sudden removal of a worker’s platform access to riders, wages and daily work, often via AI or algorithmic decisions with little explanation or meaningful recourse. The paper called deactivation one of the most severe forms of algorithmic control and said it often devastates workers’ financial stability.

This highlights a key difference between traditional employment problems and platform-based risk. In a conventional workplace, there may be clearer procedures, notice periods, HR involvement or routes to challenge decisions. In platform work, access to income can be switched off at speed, sometimes automatically and sometimes with limited transparency. For workers with no emergency fund, no specialist insurance and no reliable alternative income source, that can create an immediate household crisis.

It also helps explain why the debate is no longer just about hourly rates. If a worker’s earnings depend on opaque systems that can downgrade visibility, reduce task flow or deactivate access altogether, then the real issue is the reliability of the income stream itself. That is why modern protection discussions increasingly focus on process, appeals, data rights and human oversight alongside pay.

Why policymakers are focusing on status, benefits and digital rights

Governments and international bodies are increasingly recognising that social protection systems were not designed for all forms of digitally mediated work. The ILO’s Global Policy Tracker on Digital Labour Platforms shows that jurisdictions are now regulating employment status, remuneration, automated decision-making systems, working time, workers’ personal data, occupational safety and health, contract termination and collective bargaining. That breadth reflects a growing understanding that the challenge is structural, not narrow.

In Europe, lawmakers have been explicit about the connection between platform regulation and social benefits. The European Commission said the platform-work directive aims to ensure people working through platforms can fully enjoy the labour rights and social benefits they are entitled to. That is a notable shift because it acknowledges that the future of work is not only about innovation or convenience, but about whether people can still access basic protections when work is organised through apps and algorithms.

The wider global trend points in the same direction. The OECD’s 2025 social-protection report found that across 11 of 19 industries, between 30% and 40% of workers thought negative impacts from robots were likely over the next five years, and it said a similar pattern appears for replacement by AI and by a person providing a similar service on an internet platform. An ILO-backed discussion under South Africa’s G20 presidency also argued that rapid technological developments are reshaping both work and social protection, meaning governments need to modernise systems proactively rather than react after gaps widen further.

What this means for UK households trying to protect income and wellbeing

For UK residents, the exact legal and financial position will depend on employment status, tax setup, existing workplace benefits and the type of work being done. But the broader lesson is practical: if part or all of your income comes from freelance, contract or platform work, it is worth checking what support would actually be available if you could not work because of illness, poor mental health, injury or sudden loss of access to assignments. Many people assume they have more protection than they really do.

This is especially important because mental-health-related disruption can be financially significant even when it does not involve a formal long-term absence. Lower concentration, burnout, anxiety, sleep disruption and stress-related pauses in work can all reduce earnings for self-employed and gig workers who are effectively paid only when they are active and available. A work pattern shaped by constant monitoring, uncertain pay and low control can therefore hit both wellbeing and cash flow at the same time.

That does not mean every gig worker is unprotected, nor that technology only creates harm. The ILO Director-General said in 2025 that technology should be harnessed to create safer, healthier workplaces while ensuring innovation protects workers’ rights and wellbeing. But it does mean individuals should be realistic about the gap between flexible earning and genuine security. Reviewing emergency savings, understanding what state support may apply, checking whether any insurance or protection options fit your circumstances, and seeking regulated guidance where needed are all sensible steps when income depends on volatile systems.

The evidence from 2025 and 2026 points to a clear three-way squeeze on gig workers: weak social protection, growing mental-health risks linked to insecurity and surveillance, and AI systems that can influence pay, task allocation and deactivation without much human recourse. In that context, the old assumption that work naturally comes with a safety net looks increasingly unreliable. For many workers, the burden of managing risk is shifting onto the individual, often without the income stability needed to do it comfortably.

That is why this discussion matters beyond the gig economy itself. As algorithmic management spreads into more sectors, the line between platform work and mainstream work may become less distinct. The deeper question is not simply whether gig workers are left without a safety net, but whether protection systems can keep pace with a labour market shaped by data, automation and mental-health pressures. For households trying to plan responsibly, the most helpful starting point is clarity: know where your cover begins, where it ends, and where a gap could leave you exposed.