HOW CAN WE BUILD INTELLIGENCE SERVICES THAT ARE BOTH EFFECTIVE AND ACCOUNTABLE?

How can we build intelligence services that are both effective and accountable?

When Edward Snowden left that NSA office in Hawaii with a thumb drive full of secret files, he didn't just reveal government spying - he revealed a contradiction in our democratic government. How do you grant intelligence agencies the authority they require to safeguard citizens while also making sure they never evolve into the very danger citizens require protection from?

John Locke cautioned that absolute power corrupts absolutely, even if exercised with the best of intentions. But Thomas Hobbes taught us that in the absence of competent security, we invite a return to the state of nature in which life is "solitary, poor, nasty, brutish, and short.". Most modern democracies have attempted to balance this dilemma through parliamentary committees, inspector generals, and judicial warrants. The traditional models of oversight were crafted for a less complicated world - a world in which spying was done by human agents and wiretaps, not by algorithmic monitoring of billions of digital communications.

I think we require something much more fundamental: an intelligence system constructed from its foundations on principles of accountability, expertise, and democratic legitimacy. Not one more oversight committee, but a total rethinking of how democratic societies can even employ secret power without losing their souls.

The Current System is a mess

Before offering solutions, we must grasp why current oversight mechanisms so routinely fail. The pattern is drearily predictable: intelligence agencies act secretly, oversight bodies have no real-time access or technical knowledge, politicians rubber-stamp operations or micromanage on the basis of electoral politics instead of security requirements.

Take the CIA's post-9/11 torture program. For all the layers of oversight in place - congressional intelligence committees, inspectors general, legal counsel - the program ran for years using methods that plainly breached international law and American values. When oversight finally took effect, it was reactive, not preventative. By that time, the reputational damage to America's moral authority was irreversible. Or take the NSA's bulk metadata collection. This was a program that gathered information on nearly every American's communications, justified by secret legal interpretations that pushed statutory language to fantastical lengths. The secret FISA court that was meant to oversee the program became, in fact, a rubber stamp - approving more than 99% of government applications.

This raises the obvious question: what if it were possible to devise a system that maintains operational effectiveness while building accountability at all levels? That's the problem I posed to myself, using concepts from political philosophy.

At the heart of this new model sits a single, centralised Ministry of Intelligence and National Security. Unlike the alphabet soup of competing agencies that characterises most modern intelligence communities, this ministry would consolidate all intelligence functions under unified leadership. But consolidation isn't the innovation -accountability is.

The Ministerial Directors: Expertise Without Politics

The first revolutionary element is a panel of three Ministerial Directors (MDs) who review every operational request. But here's the crucial difference: these aren't political appointees chosen by presidents or prime ministers. They're selected by an independent Ministerial Conduct Authority (MCA) based purely on expertise, experience, and ethical integrity.

This addresses what political theorist Joseph Schumpeter identified as democracy's core limitation: citizens can choose leaders, but they can't directly evaluate every policy decision. In intelligence work, this problem is acute -voters can't meaningfully assess the merits of a signals intelligence operation against foreign adversaries. By embedding professional expertise at the decision-making level, we solve the knowledge problem while maintaining democratic legitimacy through the broader institutional framework.

The MDs must approve any proposed operation by at least a two-thirds majority. No single director can authorize operations unilaterally. This quorum system, inspired by ancient Roman checks on executive power, ensures deliberation and prevents the concentration of authority that Lord Acton warned us about.

But the MDs aren't just gatekeepers - they're moral reasoners. Drawing on Isaiah Berlin's concept of positive liberty, they must evaluate whether operations actually enhance citizens' freedom to live secure, autonomous lives, not just whether they're technically legal. They apply utilitarian calculus in the tradition of Jeremy Bentham, weighing potential benefits against probable harms, but within constraints derived from John Rawls' theory of justice - asking what intelligence powers citizens would accept from behind the "veil of ignorance," not knowing whether they might someday be targets themselves.

Political Oversight Without Operational Interference

Once the MDs approve an operation, it proceeds to the political layer - either the Prime Minister or a designated Minister for Intelligence and National Security. But here's the key: politicians can only approve or reject proposals, never modify them. They cannot alter scope, change targets, or influence operational methods.

This separation preserves democratic legitimacy while preventing the kind of political interference that has plagued intelligence services from the Bay of Pigs to the Iraq WMD intelligence. Politicians provide the democratic consent that Locke argued was necessary for any exercise of government power, but they cannot substitute their amateur judgment for professional expertise on operational details.

It's a solution that acknowledges what Jürgen Habermas identified as the tension between democratic legitimacy and technocratic effectiveness. In complex modern societies, meaningful democratic deliberation about every policy detail becomes impossible. Instead, democracy operates through the selection of representatives who exercise bounded discretion within institutional constraints designed to protect fundamental rights.

Real-Time Judicial Integration

For operations that intrude on citizens' rights - communications interception, physical surveillance, covert device access - the system requires judicial authorisation. But unlike traditional warrant processes that can take days or weeks, judges are integrated into the operational system through a secure "judicial corridor."

This isn't some distant courthouse where lawyers file papers and wait for responses. It's real-time constitutional protection. Intelligence officers can contact judges instantly via secure channels, presenting coded briefings that convey essential information in under thirty seconds. Emergency verbal authorizations are immediately logged in immutable records, creating a permanent audit trail.

The inspiration comes from Michel Foucault's insight that modern power operates through surveillance and documentation rather than dramatic public displays. By making every authorisation immediately visible and permanently recorded, the system creates what Foucault might call a "panopticon of accountability" - intelligence officers know they're always being watched, even in the most sensitive operations.

Emergency powers exist but are tightly constrained. They can only be invoked for imminent threats to national security, last a maximum of 24 hours, and can be terminated immediately by seven judges acting together. The specific number and duration might seem arbitrary, but they reflect Carl Schmitt's warning about states of exception - extraordinary powers must have clear boundaries, or they risk becoming the new normal.

The Ministerial Conduct Authority: Watching the Watchers

The MCA functions as the apex accountability mechanism, monitoring both the Ministry and the MDs. Its staff -recruited from the civil service through open competition - have full access to operational data, emergency logs, and authorisation records. But here's the crucial innovation: they're incentivised to find genuine problems. Career advancement within the MCA depends on uncovering actual misconduct, not maintaining cozy relationships with the intelligence community.

Meanwhile, the Ministry monitors the MCA, creating mutual oversight that eliminates single points of failure. Neither organization can go rogue without the other immediately detecting it. It's inspired by James Madison's insight that "ambition must be made to counter ambition" - but applied with modern institutional sophistication.

Parliamentary intelligence committees retain oversight of the MCA, ensuring ultimate democratic accountability. But they don't have access to current operational details, preserving the compartmentalization that effective intelligence work requires. This addresses the legitimate concern, raised by theorists like Leo Strauss, that some forms of knowledge may be too dangerous for public consumption while still maintaining meaningful democratic control.

Technological Safeguards and Philosophical Foundations

All operational requests, approvals, authorisations, and communications are stored in append-only, air-gapped systems. Hardware Security Modules enforce two-person authorisation for sensitive decisions. No record can ever be deleted or altered - every operational action leaves an indelible digital trace.

This reflects what John Stuart Mill argued in "On Liberty": that government power must be exercised transparently and remain subject to public scrutiny, even when that scrutiny is delayed. The technological infrastructure ensures that even classified operations will eventually be subject to historical review and democratic judgment.

The ethical foundations draw explicitly on liberal political theory. From Locke comes the principle that legitimate authority derives from consent and must operate within legal constraints. From Bentham comes the utilitarian framework for weighing competing interests and calculating proportional responses. From Rawls comes the commitment to protecting individual rights even when collective security is at stake.

But the system also incorporates what Amartya Sen calls the "capability approach" - focusing not just on preventing harm but on preserving citizens' ability to flourish as autonomous agents. Intelligence operations must consider whether they're protecting or undermining the very values they claim to defend.

The Democratic Deficit Objection

Critics might argue that this system lacks democratic accountability - that secret operations by unelected officials, however well-constrained, violate the fundamental principle that democratic citizens should control their government's actions.

This objection deserves serious consideration. Robert Dahl's work on democratic theory emphasizes that meaningful democracy requires informed citizen participation in decisions that affect their lives. How can citizens meaningfully participate in decisions they know nothing about?

The answer lies in what John Rawls called "democratic proceduralism." Citizens don't need to approve every specific action if they can approve the institutional framework that constrains and channels official power. Democratic legitimacy comes from the people's representatives establishing the system, defining its boundaries, and retaining the power to modify or abolish it.

Moreover, as political theorist Dennis Thompson argues in "Democratic Secrecy," some forms of secrecy actually enhance democratic values by protecting the privacy and security that make free deliberation possible. The question isn't whether intelligence work can be completely transparent - it can't - but whether democratic institutions can effectively control secret power.

The Effectiveness Challenge

Intelligence professionals might argue that this multi-layered system is too cumbersome for effective operations. In the real world, they might say, threats don't wait for committee approvals and judicial review.

But this objection misunderstands how the system would operate in practice. For routine intelligence gathering - the bulk of what intelligence services actually do - the process would be straightforward and rapid. For emergency situations, the system provides immediate authorization mechanisms that are actually faster than most existing bureaucratic processes.

More fundamentally, the objection assumes that effectiveness and accountability are fundamentally in tension. But history suggests the opposite. The intelligence failures of the past two decades - from 9/11 to Iraq WMDs to Russian election interference - often resulted from too little oversight, not too much. Groupthink, confirmation bias, and institutional momentum led to spectacular failures that proper oversight might have prevented.

As philosopher Karl Popper argued, institutions work best when they're designed to catch and correct errors quickly. The proposed system creates multiple opportunities for course correction before small mistakes become large disasters.

The International Relations Problem

Modern intelligence work is increasingly international. How would this domestic accountability framework handle operations conducted with foreign partners who have different oversight standards? What about operations on foreign soil where domestic legal constraints might not apply?

This is where the philosophical foundations become crucial. Drawing on the cosmopolitan tradition in ethics - from Immanuel Kant to Peter Singer - we might argue that moral constraints on government power don't stop at national borders. If bulk surveillance is wrong when conducted against American citizens, is it acceptable when conducted against German or Japanese citizens?

The practical answer is more complex. The system would maintain different standards for operations against foreign nationals, as international law permits. But even these operations would require MD approval, judicial review for actions affecting citizens of allied democracies, and MCA oversight for all activities. The goal isn't to hamstring legitimate intelligence cooperation but to ensure that international operations meet the same standards of proportionality and legal justification that domestic operations require.

The Cultural Context Question

This system assumes a particular political culture - one with strong rule-of-law institutions, professional civil service traditions, and democratic norms. Would it work in societies without these prerequisites?

Here, we might draw on the work of political scientist Francis Fukuyama, who emphasises that effective institutions require not just good design but appropriate cultural foundations. The proposed system would likely work best in established liberal democracies with strong constitutional traditions.

But that's not necessarily a limitation. As philosopher John Rawls argued, political theories need not be universally applicable to be valuable. If this framework can improve intelligence governance in democratic societies, that's worthwhile even if it can't solve the problems of authoritarian regimes.

Moreover, institutional design can help shape political culture over time. By embedding accountability mechanisms into the structure of intelligence work, the system might help strengthen democratic norms and practices more broadly.

The Church Committee's Lessons - Learning from past failures

In the 1970s, Senator Frank Church's committee exposed decades of intelligence abuses - domestic surveillance of civil rights leaders, assassination plots against foreign leaders, illegal experiments on unwitting subjects. The committee's work led to important reforms, including the establishment of congressional intelligence committees and the Foreign Intelligence Surveillance Court.

But these reforms ultimately proved inadequate, as the post-9/11 revelations demonstrated. Why? Because they focused on adding oversight layers to existing systems rather than redesigning the systems themselves. They treated accountability as an external constraint rather than an internal feature of institutional design.

The proposed system learns from this history by building accountability into the decision-making process itself. Rather than hoping that external monitors will catch problems after they occur, it prevents problems by requiring multiple authorisations before any sensitive action can proceed.

The Technology Question

Modern intelligence work increasingly relies on artificial intelligence, machine learning, and algorithmic analysis. How would the proposed oversight system handle these technological capabilities?

This is where the philosophical foundations become especially important. If we accept John Stuart Mill's harm principle - that government action is justified only to prevent harm to others - then algorithmic surveillance tools must be evaluated based on their potential for both preventing and causing harm.

The MDs would need technical expertise to evaluate AI-powered intelligence tools, asking questions like: Does this algorithm exhibit bias against particular groups? What are the false positive and false negative rates? How transparent is the decision-making process? Can individuals challenge algorithmic determinations that affect them?

Drawing on recent work in AI ethics by philosophers like Luciano Floridi and Cathy O'Neil, the system would need to ensure that technological capabilities serve human values rather than replacing human judgment entirely.

Foreign Operations and Global Justice

One of the most challenging aspects of modern intelligence work involves operations conducted outside national borders. How should democratic societies constrain their intelligence services when operating abroad?

Traditional approaches to this question have relied on legal formalism - domestic laws apply domestically, international law applies internationally. But this creates obvious opportunities for abuse, as intelligence services can simply conduct abroad what they're prohibited from doing at home.

The philosophical challenge is deeper. Political theorist Thomas Pogge argues that principles of global justice require wealthy democracies to consider how their actions affect people in other countries, not just their own citizens. Philosopher Peter Singer's utilitarian approach would suggest that nationality is morally irrelevant—harm to foreigners counts just as much as harm to citizens.

The practical solution might be a tiered system. Operations against citizens of allied democracies would require judicial review, recognizing the shared values and reciprocal relationships that characterize democratic alliances. Operations against nationals of authoritarian regimes might require only MD approval, reflecting the unfortunate reality that reciprocal constraint isn't always possible. But even these operations would remain subject to proportionality requirements and MCA oversight.

This approach draws on the just war tradition, particularly Michael Walzer's work on the ethics of warfare. Even when operating against hostile powers, democratic societies must maintain moral constraints that distinguish them from their adversaries.

The Question of Democratic Secrecy

Perhaps the deepest philosophical challenge involves what Dennis Thompson calls the "democratic dilemma of secrecy." Democratic theory assumes that citizens can hold their representatives accountable for their actions. But how can citizens evaluate representatives' decisions about matters they know nothing about?

This dilemma has no perfect solution, but it's not unique to intelligence. Citizens can't meaningfully evaluate most technical policy decisions - from nuclear power plant safety to pharmaceutical regulation to climate change mitigation. Democratic theory has evolved to accommodate what John Stuart Mill called the "competence principle" - the idea that complex decisions often require specialised expertise that ordinary citizens lack.

The solution isn't to abandon democratic control but to structure it appropriately. Citizens can establish institutional frameworks, select representatives, and periodically evaluate results even when they can't monitor every decision. Democratic accountability becomes a matter of institutional design rather than direct oversight.

Moreover, as Jürgen Habermas argues, democratic legitimacy doesn't require that every citizen understand every government action. It requires that government actions be justifiable in terms that citizens could accept if they had the relevant information and engaged in rational deliberation.

The proposed system attempts to institutionalise this kind of rational deliberation through the MD review process while maintaining ultimate democratic control through parliamentary oversight and political approval requirements.

Conclusion

The tension between security and liberty will never be fully resolved. Every generation of democratic citizens must strike this balance anew, responding to new threats while preserving the values that make democratic life worthwhile. But we shouldn't need to ever resign ourselves to the false choice between ineffective oversight and unaccountable power. We can build intelligence services that are both effective and accountable - but that choice is ours.

OPED

Author: Deni Darenberg

Checked by PoliticsWeekly Editors: Yes

Key Topic: National Security

Author Previous Works: (a) Macroprudential Regulation in the US and Canada following the 2008 financial crisis (b) Through the lens of derivatives, has market sovereignty undermined fiscal sovereignty

Protected Work

To prevent the unlawful reproduction of this protected work, your ability to copy/paste/edit/reproduce this original work is restricted.

See other stories →