Human Agency Is the Next Responsible Investing Frontier

Executive summary: A growing share of modern business models now treats human vulnerability as a source of revenue. Using data, behavioural design, pricing systems, legal asymmetry, and artificial intelligence, firms increasingly steer consumers and small businesses toward outcomes they would be unlikely to choose under conditions of clarity, symmetry, and genuine control. This is not traditional commerce. It is an emerging model of agency degradation: a business logic that depends on weakening the ability of people and smaller firms to understand, choose, refuse, leave, or retain control.

This pattern appears across luxury, media, platforms, software, mobility, and digital goods. It includes dependency on intermediating software, opaque pricing, conditional ownership, consent mechanisms that secure compliance without understanding, and exit friction that makes departure costly or impractical. Regulation will remain important, but it is too slow to be the only response. Responsible investment should therefore identify agency-degrading business models early, price that risk, disengage capital and institutional legitimacy from them, and support firms that create value without reducing the autonomy of the people and businesses they serve.

Commerce is fine - an agency destruction is not

An agency-degrading business model is one that systematically profits by reducing the ability of customers or counterparties to make informed, voluntary, and reversible decisions. Such a model is a growing part of mainstream business and treats human weakness as a commercial asset: something to be identified, shaped, scaled, and monetised. The pattern appears in different guises. In luxury and identity markets, insecurity and status anxiety become margin. In media, outrage and emotional activation displace understanding. In data platforms, users become behavioural inventory. In software services, small businesses are enclosed inside recurring tollbooths. In cars, devices, subscriptions, and digital products, ownership becomes conditional access, pricing becomes harder to understand, and customers collectively fund complexity they may not want.

These are not separate irritations. They are signs of a wider shift in the business model. Companies are using psychology, big data, interface design, pricing control, legal asymmetry, and increasingly artificial intelligence to identify vulnerabilities, shape behaviour, and monetise the result. The issue is not commerce itself. Commerce has always been competitive, self-interested, inventive, and often brutal. The issue is a commercial model that treats that brutality as permission to confuse, profile, trap, exhaust, and progressively disarm the weaker party in the transaction.

This is where responsible investing has to move next. Environmental standards, labour standards, governance tests, and supply-chain scrutiny remain necessary, but they do not capture the full social effect of modern business. A company can look clean, efficient, innovative, and well governed while operating a model that degrades trust, consent, ownership, competition, attention, autonomy, and practical choice. Ethical capital should not wait for every manipulation to be named by legislation and tested through years of enforcement. It can identify agency-degrading business models now, price the risk, withhold legitimacy, and support businesses that create value without weakening the people and firms around them.

Consumer protection is for asymmetry

The difference between legitimate commerce and agency-degrading business is already recognised, indirectly, in the history of consumer protection. Business-to-business relationships are still largely governed by contract, bargaining power, and the assumption that commercial parties can look after themselves. Consumer law developed because that assumption does not hold when the buyer lacks information, legal leverage, technical knowledge, time, or practical exit.

Rules against misleading conduct, cooling-off periods, cancellation rights, product safety requirements, and protections against unfair commercial practices were not decorative reforms. They were responses to real asymmetries. They recognised that formal consent is not always real consent, and that a market exchange is not always fair merely because a transaction occurred.

The problem is that business models are now changing faster than consumer protection can adapt. The same asymmetries are deepening, spreading, and being designed into products, platforms, pricing systems, media models, small-business infrastructure, and services that carry social importance beyond ordinary consumption. The concern becomes more serious when the same model spreads into communication, education, health cover, financial access, information, mobility, small-business infrastructure, and democratic debate. These are not just markets. They are systems through which people understand choices, manage risk, participate in society, and maintain practical autonomy.

Incremental control as a scalable business strategy

The danger is not merely that these practices accumulate gradually. The more important point is that gradual movement is now part of the model. Digital environments allow companies to move customers, users, viewers, drivers, subscribers, and small businesses one step at a time towards preferred commercial outcomes. Each step can be made small enough to seem tolerable, technical, reversible, or barely worth contesting. The cumulative effect is a steady transfer of power from the weaker party to the company controlling the system.

This is where scale and information technology change the nature of the problem. A physical shop could mislead, overcharge, pressure, or obscure, but it could not continuously test every customer, alter interfaces in real time, personalise friction, automate retention, vary offers, measure abandonment points, and optimise the whole environment around commercial conversion. An always-on, algorithmic environment can do exactly that. It can turn gradualness into a strategy of control.

The rising tide has not lifted all boats. It has gradually raised the waterline around consumers, small businesses, and institutions that lack the leverage to resist each individual change. One subscription, one hidden fee, one locked feature, one cancellation barrier, one platform dependency, one artificial scarcity tactic, one pricing opacity, one forced interface change, and one conditional access rule at a time, the terms of exchange are rewritten. Each change can be defended as marginal, legal, efficient, or normal. Taken together, they describe a market in which practical agency is being steadily weakened.

Sector signals of a broader model shift

Luxury and identity markets show how commercial value can move from serving desire to cultivating vulnerability. This requires care, because identity signalling is not illegitimate. Luxury is not inherently corrupt. Humans have always used objects, clothes, spaces, rituals, images, brands, and experiences to express status, belonging, taste, aspiration, beauty, pleasure, and self-conception. A serious argument cannot reduce all non-material consumption to manipulation.

The issue is not whether a product carries symbolic value. The issue is whether the business serves an existing human desire or manufactures the vulnerability on which its revenue depends. A luxury product can be well made, durable, beautiful, culturally meaningful, repairable, and fairly priced within its own symbolic economy. The concern begins when the product becomes secondary to the engineered pressure around it: insecurity, exclusion, artificial scarcity, social comparison, influencer dependence, body anxiety, status anxiety, and the fear of being diminished. At that point, the business is no longer simply attaching meaning to value. It is manufacturing psychological need and attaching a product to it.

Media shows a related shift from informing to activating. The issue is not simply bias, partisanship, or low standards, which are old problems. The newer issue is the monetisation of psychological activation as an operating model. Much of the media economy no longer rewards understanding. It rewards attention, and attention is most reliably captured by fear, outrage, grievance, anticipation, tribal identity, unresolved tension, and addictive incompleteness. A media business may still describe itself as informing the public, while its commercial engine rewards keeping people emotionally activated rather than better informed.

Big data platforms are the clearest technical expression of the same logic. Many do not simply serve customers. They observe, profile, rank, nudge, price, retain, distract, and monetise them. The person using the service may appear to be the customer, but the revenue model often treats that person as inventory: an advertising surface, behavioural dataset, engagement unit, training input, or prediction object. The problem is not data itself. Data can improve services, reduce friction, reveal unmet needs, widen access, support better decisions, and make products more useful. The issue is the commercialisation of data through an extractive behavioural model.

That model often depends on opacity. The user does not know what is collected, how they are profiled, how prices are shaped, why certain options are shown, why other options disappear, or how much of their experience is designed around their interests rather than the company’s monetisation strategy. Consent then becomes theatrical. The user has clicked, accepted, continued, agreed, upgraded, subscribed, renewed, or failed to opt out. Legally, that may be enough. Socially, it is often fiction. A process designed to secure compliance without understanding is not meaningful consent. It is administrative camouflage.

The operating tools of agency degradation

The sector examples matter because they show the breadth of the shift. The operating tools show how it is implemented. These tools are not confined to one industry. They move across sectors because they are commercially attractive, professionally packaged, and increasingly normalised by the companies with enough scale to set market expectations.

• Dependency as infrastructure control

The first tool is dependency architecture. Small firms increasingly depend on software systems for payments, accounting, booking, logistics, marketing, customer management, online visibility, compliance, stock control, delivery, and communication. Many of these tools are useful. The problem is not software. The problem is the conversion of business infrastructure into a tollbooth economy, where a small business may no longer own its route to market, its customer relationship, its payment flow, its data, its discoverability, or even its basic operating systems.

This mirrors the consumer experience, but with wider economic consequences. Small businesses are not merely customers. They are productive economic actors. If they become dependent on software intermediaries they cannot meaningfully negotiate with or leave, the result is a weakening of commercial agency. Constructive software increases the capacity of small firms to operate, compete, and retain value. Extractive software captures workflows, customers, payments, visibility, and data, then charges recurring rent for continued access to commercial functionality. This is not a complaint about subscription fees. It is a question of whether the economy is being made more productive or merely more intermediated.

• Pricing as permission architecture

The second tool is pricing architecture. The old concern was planned obsolescence: the suspicion that products were designed to fail earlier than necessary. The newer model is often subtler. Products do not have to fail. They can be designed so that customers collectively fund complexity they may not want, while manufacturers capture high-margin revenue from selective activation.

Cars are one example. Software is another. Modern products are increasingly built around common platforms that include higher-spec capacity from the outset. This can make sense from a manufacturing point of view. Luxury features can be produced at high volume and lower unit cost. Product variation can be reduced. Upgrade paths become easier to manage. The ethical issue lies in the pricing structure. If additional capacity is already built into the base product, then the cost of that capacity is likely to be spread across the whole customer base. Customers who do not want or use the feature still help fund the more complex platform. Those who later pay to activate the feature may then be paying not for new production value, but for permission to access capacity that already exists.

The base price funds the platform, while the upgrade price monetises permission. This is not always illegitimate. Manufacturing efficiency can reduce costs. Standardised platforms can improve reliability. Optional upgrades can allow flexibility. The concern arises when the customer is not told clearly what they are funding, what they own, what is merely disabled, and how much of the upgrade price reflects real additional cost rather than control over access. The wider pattern is the same: the company uses design, information, pricing, and control to move value away from the customer and towards the producer after the point of sale.

• Ownership as conditional access

The third tool is ownership erosion. Customers increasingly buy products that behave less like property and more like conditional access. They purchase devices, vehicles, software, appliances, media, tools, and services whose functionality can later be changed, limited, withdrawn, updated, re-priced, or locked behind new terms. The customer pays, but does not fully control. They may not be able to repair, resell, modify, downgrade, maintain an older interface, refuse an update, preserve a simpler version, or continue using the product without ongoing consent to revised conditions.

This is economically significant. Ownership has historically been one of the ways individuals and small businesses protect themselves from dependency. When ownership becomes conditional, revisable, and technically mediated, the customer’s bargaining position weakens. A responsible investment framework should therefore treat ownership degradation as a serious indicator. A company that sells products while retaining practical control over their use is not simply innovating. It is changing the relationship between commerce and autonomy.

• Consent as procedural fiction

The fourth tool is consent dilution. The legal form of consent is often preserved, while the substance is weakened. A customer can be deemed to have accepted terms they did not read, understood implications they could not reasonably assess, consented to data uses they could not map, accepted pricing changes hidden in communications they did not notice, or agreed to product changes they had no practical ability to refuse. This is not a marginal defect in the process. It is often central to the process. The company needs consent to be legally effective, but not necessarily meaningfully informed.

• Exit as engineered friction

The fifth tool is exit friction. Many models depend less on winning continued preference than on making departure inconvenient, confusing, costly, risky, or incomplete. Cancellation is harder than sign-up. Data export is limited. Compatible alternatives are scarce. Customer histories, contacts, rankings, reviews, and workflows are trapped inside the platform. A consumer loses accumulated convenience. A small business risks losing access to customers, payments, records, or visibility. The company then describes retention as loyalty, although some of it is merely captivity by design.

When non-compliance becomes a priced strategy

These operating tools are strengthened by a familiar legal and compliance pattern. The issue is not occasional error. It is the strategy of acting first in the company’s favour, banking the gains, and treating later enforcement as a manageable cost of doing business.

The sequence is familiar. Act first, shift the burden onto customers, suppliers, small businesses, or regulators, and see whether anyone has the power, time, money, or legal standing to challenge it. Litigate, delay, lobby, fragment the affected parties, settle years later if necessary, pay a fine smaller than the profits already secured, avoid restoring the original position, issue a carefully managed apology, and move on with the structure largely intact. That is not compliance failure. It is profitable non-compliance.

The fine is then presented as accountability. In practice, it may function as a retrospective licence fee, paid after the advantage has been captured and without disentangling the people or businesses caught inside the system. This should be a core indicator of extractive intent. The question is not only whether a company has been fined. The sharper questions are whether it reversed the harm, restored customers or counterparties to the position they would have occupied, disgorged the profits, changed the interface, altered the contract, improved exit, disciplined management, and stopped repeating the behaviour across jurisdictions or product lines.

Where a company can mislead, profit, delay, settle cheaply, keep the customers, and continue operating the same model, investors should not treat the matter as closed. They should treat it as evidence of business-model intent. This is especially important because the companies using these models are not marginal or disreputable outliers. The practices are flowing outward from an oligopolistic core of large, data-rich, professionally managed companies whose methods set expectations across the economy.

This is how business ethics changes. Not usually through a formal declaration, but through repeated commercial success. Practices once considered sharp, dubious, or reputationally risky become standard. The language changes first. Manipulation becomes personalisation. Lock-in becomes ecosystem. Price discrimination becomes optimisation. Surveillance becomes insight. Inertia becomes retention. Forced complexity becomes innovation. Strategic non-compliance becomes regulatory uncertainty. The economy then absorbs the new norm.

Why capital must move ahead of regulation

Law matters, but law is slow. Enforcement is slow. Legislative language is slow. Litigation is slow. Companies with scale, data, lawyers, cash flow, and political reach can often profit during the delay. Consumer protection has won important victories, but those victories usually arrive after the business model has already moved on.

Responsible investment should not wait for every manipulation to be named in legislation, tested in court, appealed, delayed, and eventually settled. Capital can move earlier. It can identify business models that depend on agency degradation, price that risk, withhold legitimacy, alter procurement standards, and support competitors that make money by serving rather than manipulating.

This is not an argument for investment to replace law. It is an argument for investment to stop hiding behind the slowest possible mechanism of change. Where a business model depends on exploiting weakness, obscuring value, degrading ownership, diluting consent, creating dependency, or delaying accountability, investors do not need to wait until the final legal appeal has ended before recognising the ethical direction of travel.

The investment test: does the model strengthen or weaken agency?

The next responsible investing frontier is human agency. The test should be direct: does this company strengthen or weaken the ability of people and smaller businesses to understand, choose, own, leave, repair, compare, compete, refuse, and consent?

That question can be turned into practical indicators, without becoming a philosophical exercise. Investors can examine whether revenue depends materially on hidden fees, opaque pricing, manipulative defaults, renewal traps, cancellation friction, behavioural advertising, artificial scarcity, upgrade pressure, insecurity, compulsive engagement, dependency, or strategic non-compliance. They can assess whether consent is clear, specific, reversible, and proportionate, or whether it is buried in processes designed to secure acceptance without understanding.

They can also examine whether customers can leave easily, export data, cancel subscriptions, repair products, retain access, resell purchases, or avoid lock-in. They can ask whether purchase confers meaningful control, or whether the company retains practical power through licences, software locks, updates, subscriptions, parts restrictions, or revised terms. They can test whether innovation improves the underlying product or mainly improves monetisation, targeting, friction, bundling, dependency, and control.

For small-business exposure, investors can ask whether the company helps firms compete or encloses them inside software, payment, advertising, marketplace, and visibility systems they cannot realistically leave. For enforcement history, they can ask whether misconduct has been followed by restitution and structural correction, or whether the company merely paid a fine and continued. These are not abstract moral preferences. They are business-model questions. They can be assessed, scored, disclosed, compared, and used to guide capital.

From diagnosis, to disengagement, to change

The first task is to name the pattern. The language is currently fragmented. People talk separately about dark patterns, addictive media, privacy, subscription traps, degraded ownership, luxury manipulation, software lock-in, opaque pricing, customer fatigue, and weak regulation. The common mechanism is agency degradation: the systematic weakening of the ability to understand, choose, own, refuse, leave, compete, and consent.

The second task is to measure it. Investment analysis already measures exposure to carbon, labour risk, governance weakness, supply-chain abuse, regulatory risk, and reputational harm. It can also measure exposure to manipulation, dependency, opacity, ownership degradation, consent dilution, exit friction, and strategic non-compliance.

The third task is to disengage. This does not mean naïve ethical consumption or performative boycotts. Individual consumer choice is too weak to carry the burden alone. The more serious levers are institutional capital, pension funds, procurement rules, university endowments, public-sector purchasing, philanthropic capital, development finance, and asset managers claiming social purpose.

The fourth task is to change incentives. Companies that build around human agency should receive capital, legitimacy, procurement preference, and reputational advantage. Companies that build around behavioural raw material should face higher capital costs, stronger disclosure demands, shareholder pressure, procurement exclusion, regulatory scrutiny, and reputational resistance. The point is not to abolish profit. It is to distinguish between profit earned by serving people and profit extracted by weakening their ability to judge, refuse, own, leave, or consent.

Conclusion: Agency is the next frontier of responsible investment

The degradation of modern business is not best understood as a collection of consumer frustrations. It is a shift in the way powerful companies think about people, small businesses, and society. Too much of the economy is moving from service to manipulation, from product improvement to behavioural engineering, from ownership to conditional access, from pricing to extraction architecture, from customers to inventory, from public information to psychological activation, and from enterprise software to dependency tollbooths.

This is not inevitable. It is a choice, a business model, and an investment question. Responsible investment should not stop at the physical environment. It must also address the human environment in which markets, trust, ownership, consent, competition, and civic life operate. A company that reduces emissions while degrading human agency is not socially constructive in any serious sense.

The next stage of responsible investing should be clear. Name agency-degrading business models. Measure their dependence on manipulation. Disengage capital, procurement, and institutional legitimacy from them. Support alternatives. Push regulation towards restitution, exit rights, ownership rights, pricing transparency, and meaningful consent.

The central question is not whether a company looks modern, innovative, efficient, or sustainable. The central question is what its business model requires from the people and smaller businesses around it. A constructive model helps them understand, choose, own, compare, refuse, leave, and participate. An extractive model makes money when those capacities are weakened. That is where responsible investing has to move next.