Slow Horses Season 5 Review: When MI5 Becomes Predictable

Official promotional artwork for Slow Horses Season 5. Apple TV+ promotional material, used for review and commentary.

Season five of Slow Horses begins with events that seem to require different explanations: public violence, political agitation, technical disruption and Roddy Ho’s implausibly glamorous new relationship. The danger lies in the possibility that MI5’s separate responses are themselves part of the design.

Adapted from London Rules, the fifth novel in Mick Herron’s Slough House series, Apple TV+’s six-episode fifth season has one of the series’ strongest plot ideas. An adversary does not need to overpower the British security apparatus if it can predict how each part will react. Institutions designed to divide responsibilities, classify threats and protect sensitive information become vulnerable when someone understands the gaps between them.

I admired that construction more than I enjoyed its execution. The season offers reversals, concealed connections and an unusually well-motivated terrorist conspiracy, but it also makes its machinery rather visible. Once the pattern begins to emerge, the characters sometimes seem to occupy positions required by the design rather than driving events through their own mistakes and decisions.

It is a good season built around an excellent premise. Complex systems do not always fail because every component is incompetent. They can fail because each component responds rationally to the wrong version of the problem.

A Campaign Against Interpretation

Television terrorism plots usually give the audience a target to anticipate, even when the characters have not yet identified it. Someone intends to assassinate a politician, destroy a building or cause mass casualties at a recognisable event, and the investigation becomes a race to discover where the decisive attack will occur.

Season five initially refuses that shape. Its incidents differ in method, scale and apparent purpose. Each creates its own emergency and attracts a response from whichever institution seems best equipped to handle it: the police pursue one kind of threat, MI5 another, while politicians react to public fear and the pressures of a London mayoral campaign. Before the authorities can stop the wider plan, they must recognise that these separate crises belong to the same one.

The attacks therefore operate first upon interpretation. Institutional specialisation allows experts to concentrate on particular problems, but it also means that information travels through channels organised around assumptions formed before the whole situation is visible. A shooting is investigated as a shooting. A technical breach becomes a cybersecurity problem. Political unrest is treated as a matter of public order or electoral strategy.

An intelligent adversary can keep those institutions reacting locally while the larger pattern develops between their jurisdictions. The vulnerability lies less in resources than in coordination, timing and classification. Someone must decide which details matter, which departments should communicate and when a collection of anomalies has become evidence of design. Until then, every part of the system may be functioning as expected while the whole moves in the wrong direction.

This extends an argument the series has been making from the beginning. Slough House and the Park do not simply differ in competence; they operate under different constraints. The Park has personnel, surveillance capacity and formal authority, but its information is divided by hierarchy and political interest. The slow horses have almost none of those advantages, yet their marginal position sometimes allows them to notice connections that do not fit established categories.

That does not reveal them as hidden geniuses. Their suspicions remain erratic, their coordination poor and their judgement entangled with vanity, anger and personal history. Their occasional advantage is narrower: they are accustomed to beginning with problems the institution has already misunderstood.

The attackers exploit the same weakness from the opposite direction. They do not need to make MI5 irrational. They need it to remain predictable.

A More Disciplined Kind of Extremism

The comparison with the Sons of Albion plot in Season 1 shows how far the series’ treatment of political violence has moved.

The Sons of Albion were dangerous, but also theatrical, vain and operationally weak. Their extremism depended upon performance. They wanted recognition for acting on behalf of an imagined nation, and their identity was built from slogans, symbols and the fantasy that violence would make them historically important. That visibility made them easier for more experienced people to redirect, exploit and discard.

Season five’s conspiracy is quieter and more disciplined. Its members do not require the public to understand their identity while events are unfolding. Their violence is less an announcement than an instrument for applying pressure to the city and observing how its institutions react.

The difference is one of emphasis rather than a clean opposition between amateurs and professionals. The later group is still held together by human motives: resentment, loyalty, coercion, grief and the desire to give private suffering a wider meaning. Its members do not necessarily understand every part of the operation in the same way.

That internal variation is one reason the plot works better than most television terrorism stories. The attackers possess histories rather than a single generic fanaticism. The series gives their hostility an origin without turning explanation into absolution. Understanding why people joined the campaign does not make their choices defensible; it makes those choices legible.

The threat also continues the series’ interest in consequences produced by states and intelligence services long after the officials involved consider an operation closed. Season 2 treated Cold War history as operational debris. Season 3 showed MI5 trying to suppress the human cost of its own decisions. Season 4 examined intelligence expertise after it had escaped reliable institutional control.

Season five develops that thread by imagining antagonists who understand that a state can be attacked through its reactions. Public fear, political competition and institutional self-protection can amplify an event beyond its immediate physical effect.

The group remains coordinated rather than omniscient. Its plan depends upon people behaving predictably, so every improvisation or personal deviation introduces risk. The conspirators are formidable because they have prepared carefully, not because the story grants them unlimited foresight. A plan built around institutional behaviour remains convincing only while the institutions are populated by people capable of departing from the script.

Roddy Ho and the Human Perimeter

The season’s most comic plot line also contains one of its clearest security arguments. Roddy Ho has a girlfriend, and nobody at Slough House believes this development can be accepted without investigation.

The joke works because Roddy cannot imagine himself as vulnerable. His vanity is so complete that another person’s interest appears to confirm facts he already considers obvious. His colleagues see an anomaly; Roddy sees the universe correcting a long-standing oversight.

That self-assurance makes him unusually susceptible to attention. He does not need to be deceived about his technical abilities, because those abilities are real. He needs only to be encouraged to treat specialised competence as proof of universal judgement.

Roddy’s systems knowledge and access make him genuinely valuable to MI5. His understanding of other people is far less dependable. The combination is more dangerous than either quality alone. Someone with no access cannot compromise much, while a person conscious of their limitations may recognise an attempt at manipulation. Roddy possesses access without self-doubt.

This returns to Season 2’s distinction between competence and control. Someone can understand a technical system better than almost anyone around them while misreading the motives of the person sitting across the table. Expertise does not spread evenly across a personality.

The human perimeter of a secure institution therefore includes every person allowed inside it. Encryption, access controls and monitored networks cannot eliminate the possibility that someone will appeal to vanity, loneliness, resentment or ambition. The route around a technical barrier may run through the employee convinced that nobody could manipulate him.

Roddy is an exaggerated version of that problem, but not a meaningless one. His confidence is ridiculous because the audience can see the distance between his self-image and social reality. From his perspective, however, the confidence has evidence behind it. He is routinely asked to solve problems other people cannot solve. His mistake is to assume that being indispensable in one narrow domain makes him a reliable judge of every situation involving it.

The attackers do not need to defeat his intelligence. They need to direct it.

MI5 displays a related arrogance at institutional scale. The Service possesses powerful tools and assumes that this makes it sophisticated in every dimension. Season five repeatedly shows systems capable of processing data while misunderstanding the people and incentives producing it. Roddy embodies that error without any capacity for embarrassment; the Park makes the same mistake behind locked doors and official titles.

J. K. Coe and the Absence of Explanation

I still cannot decide whether I like J. K. Coe, and that uncertainty may be what makes him interesting. He spent much of Season 4 as a quiet presence at the edge of Slough House before his contribution to its ending revealed how little the others understood about him. Season five gives him more purpose within the team, but not much more emotional transparency.

Coe is not a blank slate so much as a sealed file. His behaviour is visible, but the reasoning behind it remains inaccessible.

River’s impatience, Shirley’s anger, Roddy’s vanity and Lamb’s contempt all create problems, yet they also make those characters readable. Their reactions have patterns. Lamb’s cruelty often conceals preparation or loyalty. River’s recklessness grows from his need to act before somebody else controls the outcome. Shirley’s aggression can be connected to grief, addiction and humiliation.

Coe offers fewer points of entry. He listens, observes and occasionally reaches conclusions before anyone else, but rarely explains the emotional or moral process behind them. His stillness might indicate discipline, trauma, detachment or some combination of all three.

That ambiguity makes his behaviour harder to predict than the actions of the more openly unstable slow horses. A volatile person may still advertise what is about to happen. Coe appears capable of moving from observation to action without offering anyone else the intermediate steps.

His previous conduct established that he can use lethal force without seeking institutional approval. What remains unclear is the rule governing that capacity. Was the action exceptional and driven by immediate circumstances, or did it reveal a private moral threshold he may cross again?

The series is right not to answer too quickly. Explaining Coe would risk reducing him to another damaged officer whose trauma can be converted into a neat character key. His opacity is more unsettling because the audience must infer a system from very little evidence.

There are signs of considerable damage beneath the restraint, but trauma alone does not explain what someone believes they are entitled to do. Coe may be emotionally numb, rigorously controlled or simply unwilling to expose himself to people he neither trusts nor respects.

He is also useful. His distance allows him to see patterns that more emotionally involved characters overlook. Slough House has faced versions of this problem since the first season: an officer’s damage may contribute to both their effectiveness and their danger.

Lamb’s usual advantage is that he understands what damaged people will do next. He knows River will rush ahead, Roddy will overestimate himself and Shirley may turn pain into aggression. Coe may prove harder to manage precisely because he offers so little evidence from which to predict him.

The uncertainty is not whether he has feelings. It is whether those feelings still connect his actions to rules anyone else can recognise.

Better Plot Than Execution

The season’s central design is clever enough that the series occasionally draws too much attention to it. The early incidents work because they resist immediate connection: each has its own tone, apparent motive and institutional consequence. As relationships begin to emerge, randomness acquires shape without making every event feel identical.

Once the underlying logic becomes clearer, however, some of the mystery turns procedural. Characters begin to infer what kind of event might follow from what has already occurred. The recognition raises the stakes, but it also makes the construction more visible. A campaign that initially felt unpredictable begins to resemble a sequence whose stages can be anticipated.

The reversals create a similar tension. Several are satisfying because they force the audience to reinterpret earlier scenes. Others depend upon information being withheld until the moment when its release can produce maximum surprise. We understand why the pieces fit, but do not always feel that the characters reached the connection through accumulated investigation.

The plot drives the people more than the people drive the plot.

This becomes clearest when compared with the series’ strongest seasons. Season 2 allowed conspiracy, character and institutional history to develop as one movement. River’s impulsiveness, Lamb’s Cold War experience and Taverner’s manoeuvring altered the investigation rather than merely responding to it. Season 4 gave the same structure a more personal centre by connecting the threat to River’s family.

Season five begins with a large strategic design already in motion. Its ensemble is placed inside that design, and the six-episode structure leaves limited space for every character’s response to reshape it. Roddy, Coe, Shirley, River, Lamb, Taverner, Claude Whelan, Emma Flyte, the mayoral candidates and the members of the attacking group all compete for time.

The compression keeps the season moving, but several emotional consequences remain thin. The antagonists receive enough history to avoid becoming generic, though not always enough time for their differences to acquire full dramatic force. River continues to carry the personal consequences of Season 4, yet that strand sometimes feels adjacent to the main campaign rather than woven into it.

The technological elements are also less convincing than the social and institutional ones. The broad idea that hostile actors could exploit access, fragmented responsibilities and technical dependence works well, but some of the specific mechanisms feel like narrative shortcuts. The social engineering convinces because it grows from vanity and organisational habit; the computer systems occasionally do whatever the next reversal requires.

Claude Whelan illustrates both the strength of the premise and the occasional bluntness of the execution. His instinct for self-preservation makes him predictable. He responds to emergencies partly according to their operational demands and partly according to the danger they pose to his position.

That behaviour fits the series’ London Rules: cover your arse before someone else decides it belongs in the official account. When several senior figures protect themselves at once, however, their individually understandable decisions produce precisely the confusion an adversary would hope to create.

Season five is good at showing that institutional vulnerability does not require universal incompetence. Every participant can behave plausibly within a narrow role while the collective response deteriorates. It is less successful when the plot must move those people rapidly from one inversion to the next.

I liked the season, and its terrorist conspiracy is considerably better developed than the television norm. It gives its antagonists histories, allows their strategy to exploit recognisable institutional habits and avoids reducing the threat to one spectacular target. The design nevertheless remains stronger than the dramatic flow. I could see the intelligence of the construction even when I was not entirely caught inside it.

Season five is most convincing when its adversaries do not need to outgun MI5. They need only understand the institution well enough to let it obstruct itself.