Consider two recent events. The CEO of Eiffel Tower’s managing company stepped down owing to a simple, though sensitive, administrative issue of segregating female employees during the visit of a team of monks. On the other hand, Mark Zuckerberg, the high-profile founder-CEO of Meta Platforms who commands 61 percent voting power in the tech giant, did not lose a single dollar after what the courts agreed were his company’s major and deliberate violations relating to child safety, data misuse and platform addiction. Accountability works only if applied universally.
Meta, in August, settled a set of suits by agreeing to pay four US states about $18 billion. Two months later, another US court ordered Meta to pay $25 billion to the state of New Mexico in a separate suit alleging sharp practices on data privacy. But for the magnanimity of the judge, this penalty could reportedly have been up to $260 billion.
These are not first-time violations by Meta. Previous penalties on the company and its associates for privacy violations include the 2019 Federal Trade Commission settlement of $5 billion, the 2022 Irish authority’s penalty of €405 Million, the 2023 European Union fine of €390 million, just to mention a few. So it’s a series of continuing offences of the same nature; apart from competition violation.
However, Mark and Meta are not the only beneficiaries of the new pattern of settlement with cash penalties on the companies (with or without some modification of their companies’ business practices). As if the persons-in-charge, the conscience keepers, of the companies have nothing to do with the violations. Big CEOs and other top leadership seem to be too big for jail or even for personal fines.
Here are some big examples. Through the 1998 landmark Master Settlement Agreement, Big Tobacco companies committed $206 billion to 46 US states to offset the systemic public healthcare expenditures associated with smoking-related problems. BP settled the Deepwater Horizon oil spill for $20.8 billion in 2016. Volkswagen incurred a $14.7-billion liability the same year on account of its emissions-cheating setup.
In the financial sector, Bank of America's record $16.65-billion settlement in 2014 and JPMorgan Chase's $13 billion resolution in 2013 were stemming from the proliferation of toxic mortgage-backed securities. BNP Paribas agreed to $8.97 billion forfeiture in 2014 for illicitly processing transactions for embargoed jurisdictions. Except in a few cases the CEOs/top management of those companies did not suffer personally.
The central argument for settlement is rooted in public interest and speedy justice. In tune with this, State and regulatory frameworks have leveraged unprecedented, multi-billion-dollar financial settlements as the primary mechanisms to mitigate widespread harms on people; including children, the economy and environment.
The increasing shift towards out-of-court settlements and plea bargains, is reshaping the relationship between crime and punishment. It is blurring the lines of justice by turning criminal accountability into a financial, commercial or administrative transaction. The practical logic of such light-touch approach to justice and inverted proportionality emanates from politico-economic consequences of the fall of big companies in terms of production, unemployment and the deficiencies in the investigative-judicial machinery to deal with the increasing incidence of large and complex violations.
Even the Supreme Court of India travelled this path while closing the Sterling Biotech bank fraud case in 2025, after the promoters agreed to pay a recovery package. This was despite the hitherto strong stand that settling criminal offences was a crime against society. That weight must have forced the court to emphatically add that the judgement should not be treated as a precedent. However, this case provides a real-world example of how the line between crime and punishment can blur when pragmatism intersects with criminal accountability.
The traditional ideal of criminal justice envisions a binary system: a crime is committed and proved; a proportional punishment is handed down. Even small crimes used to invoke serious punishments in some jurisdictions. Like Jean Valjean, the protagonist in Victor Hugo’s Les Misérables was awarded five years’ imprisonment by the then French judicial system for stealing one loaf of bread.
However, the modern shift towards negotiated settlements, plea bargains and deferred prosecution agreements is fundamentally challenging this paradigm in many countries. This evolution is increasingly blurring the boundary between crime and punishment, turning justice into a transaction—monetisation of justice. The message conveyed to the public is that the severity of a crime can be mitigated by the size of a corporate bank account.
This Western model risks transforming the judiciary into a marketplace where criminal liability is bargained away behind closed doors, often shielding the powerful from genuine accountability. While settlements do not entirely erase the jurisprudence, they do fundamentally alter justice delivery. Punishment is no longer a guaranteed, transparent consequence of a crime; instead, it has become a negotiable variable. A company-only-paying for settlement approach does not reduce repeated violations. Rather, it is encouraging repeat violations.
Artificial Intelligence is going to amplify this scenario many fold. The AI bosses themselves have called out loudly on the reality of some agents turning rouge and wreaking havoc with other systems. Expert explanations regarding the reasons for such a clarion call, that too from hitherto AI die-hards, vary. Many are not borrowing the ‘end of humanity’ type arguments too. However, going by even the reported incidents on rogue AI tools, the incidents will grow; in number and size, as AI becomes a stranglehold. Many of those incidents will come out as big disputes. How will those violations be punished? Punish those rogue AI? Their creators; the natural persons or the legal persons? Natural persons would say, we had warned you.
Ultimately it will be an increasing number of civil settlements; irrespective of the crime. Including, hypothetically, a nuclear disaster caused by a rogue AI tool. As long as such settlements do not affect the natural person/s responsible, the exchequers will have plenty of money coming in. Civil settlement with corporations may assuage monetary losses that can be remedied. But where harm to life is involved directly or indirectly, a monetary settlement with those companies only is truncated justice. The persons responsible should also face justice proportionately.
The world of justice should not be an inverted one. Where even small crimes of individuals get punished harshly but big corporate crimes get settled with a fine on only the company. The world of financial settlement needs a revisit.
C K G Nair | Former Director, National Institute of Securities Markets and former member, Securities Appellate Tribunal
(Views are personal)
