Techno-Logical Leakage

On TSB’s Migration Accident




20 - 22 April 2018

On Thursday 19 April 2018, UK challenger bank Trustee Savings Bank (TSB) announced that they will upgrade their systems over the weekend to a more ‘customer-centric design’, thereby making some banking services temporarily unavailable.[1] But things didn’t go according to the plans. On Sunday, when the migration was planned to complete, customers noticed that their bank accounts remained unavailable. In the end, it took TSB over six weeks to fix the majority of the faults and start running their new banking platform. In those weeks, hundreds of thousands of people couldn’t use their bank accounts, disabling them to pay for day-to-day expenses and many customers fell victim to cybercrime attacks.[2] What does the failure of banking technologies, that happens on the level of everyday life, discloses about the entanglement between banks and society? Do banking upgrades merely optimise the user experience or do they more so optimise the bank’s debt-creating machinery?


22 April - June 2018



Weeks of back-and-forth between error messages and contradictory explanations retained customers in an endless double bind. While technical clarifications may stabilise banking systems, they won’t directly address the flawed reasoning behind the failures. Machines function to fulfil desired outcomes yet the seamlessness of user interfaces conceals these exteriorised intentions, as described by Bernard Stiegler.[3] The accident, as conceptualised by Paul Virilio, manifests itself as a material malfunction but at a deeper level is driven by a failure of reason: the unrealistic reliance on technology to function flawlessly in accordance with our desired ends.[4] When the bank’s exploitative machinery malfunctions and its back-end circuitry breaks through the veil of user-friendly interfaces, the public runs into a conceptual contradiction. Error messages, however, enclose clues of the accident and become symptoms in a broader pathology. They can provide an entry into the underlying logics of banking technology to establish a techno-logical leakage. Unpacking TSB's flawed logics will happen through focusing on a single error message about a misconfigured load balancer, a type of router that was deliberately designed to optimise their new platform. The router’s role first became publicly understood as the instigator of the accident because it led to bandwidth congestion, which the bank framed as the problem. Subsequently, the router started to disclose the dispersion of responsibilities. TSB’s strategy to outsource and blame technology was strengthened by the unresponsiveness of the Financial Conduct Authority. Post-migration, when TSB’s new platform (Proteo4UK) started working, the router commenced its predominant role of unnoticeably accelerating the enticement of becoming indebted.






Webpages took forever to load and account balances didn’t tally. Former CEO of TSB Paul Pester responded on Sky News that the vast number of customers trying to access the new website was affecting their bandwidth.[5] That is, the megabits of data that can be transmitted per second. Pester decided to blame the bandwidth and shift the fault onto technology—as if the router misconfigured itself. Was it the bandwidth that failed or TSB’s techno-logics? The bank relentlessly exhausted the transmission to its limit to reach Proteo4UK as soon as possible, while ignoring the very possibility of a breakdown­. During the accident, the mobile application spurted many error messages about a load balancer, which was misconfigured according to the investigation of IT-giant IBM.[6] A load balancer is a device that balances and routes web traffic across a cluster of servers to prevent overload. A misconfigured load balancer rapidly leads overloads of data into a bottleneck. The router runs out of memory, CPU’s work energetically and overheat the router until it literally burns-out. The scalability of routing became exploited to the moment of destruction. Technology was deployed as the solution to create a frictionless world without errors. But when disorder arrived, TSB was confident that its techno-solutionist decision to throttle the bandwidth fixes all of their mistakes. The router, as a chokepoint, elucidated how the congestion of bandwidth disrupted TSB’s IT infrastructure. But the router also became surrounded by contradictory explanations behind its malfunction. These hyperlinked narratives took shape in and around the router, exceeding its mere purpose of distributing data. At first glimpse, all technical explanations appear to be linked to each other and elucidate the website’s unavailability. Closer attention reveals the inconsistency of these explanations. The messiness of banking upgrades and failures hold together the smokescreen of the bank’s capitalist intentions. Nonetheless, routing is just one technological process through which hyperlinked narratives became prevalent. The various actors involved in the migration—from regulatory bodies to tech firms—collaboratively plot their capitalist desires.


As the regulator of the financial services industry in the United Kingdom, the Financial Conduct Authority (FCA) was actively involved in the development of Proteo4UK. Despite the lack of evidence regarding the platform’s readiness, the FCA permitted the bank to start the data migration. My Freedom Of Information Act Request aimed to disclose the mismanagement of the migration. The FCA couldn’t release the information because the commercial interests outweighed the public interest test (figure 1). Commercial interests disabled the public disclosure of the responsibilities of the organisations involved in the migration. How can the FCA criticality regulate the financial market when its members are its only incomes? Who is watching the watchdog? Their funding model likely prevented the acceptance of my FOIA request because ‘disclosure would harm the reputation of the financial market’ (figure 1). Moreover, outsourcing ensures that Proteo4UK cannot be holistically scrutinised as its myriad components became de-localised. Outsourcing the assembly of banking systems fragments accountability and an understanding of the system's predatory intentions. Instead, Paul Pester became the embodiment of the problems. His public prosecution and departure from TSB thus became a fast (yet trivial) solution to regain the fabricated stability and trust in banking. Pester's dismissal will not affect the technical environment that contributed to the migration. Financial corporations and IT consultancies will continue to outsource and solve whatever tasks banks give them as they won’t be held responsible for the repercussions.  





2008 - 2018



TSB’s IT upgrade arose in the aftermath of the financial crisis of 2007-08. The magnitude of many financial institutions grew to the extent that their failure triggered a global financial disaster. These too big to fail institutions “reap profits if their trades pan out, but taxpayers can be stuck picking up the tab if their big bets sink the company."[7] Following the £20 billion taxpayer bail-out in 2008, the European Commission enforced mega-bank Lloyds Banking Group (Lloyds) to split off TSB.[8] The government also steered Lloyds to take over the bankrupted HBOS, creating an immense patchwork of banking systems.[9] Lloyds rapid divestment made it infeasible to instantly develop a new banking platform, thus TSB rashly switched to a cloned version of Lloyds’ banking systems.[10] In retrospect, inheriting Lloyds’ IT infrastructure became the precursor of TSB’s accident. Following the complete separation from Lloyds in 2013, TSB was taken over by the Spanish mega-bank Sabadell in 2015. Sabadell proposed an upgrade towards a new ‘customer-centric’ banking platform, which would cut off TSB’s millions of pounds in monthly fees for Lloyds’ IT system.[11] These costs likely propelled Sabadell’s reckless ‘big bang’ approach, migrating all data at once without a fall-back plan.[12] The Spanish Mega-Bank deliberate chose to execute codes without hard evidence of its performance outside of virtual test environments.[13] Society served as the actual test when it was too late to prevent the repercussions of flaws in the code. The fixation for quick profit is what drives the bank’s inconsiderate decisions. The aftermath of the crisis produced dependencies on labyrinths of banking systems that consistently empower banks to act on their terms.





After the major IT catastrophe, TSB lost billions of pounds to repair their systems and yet has to ‘fix’ 400 issues, which won’t be solved until the end of 2019 (at the earliest) says their new CEO Debbie Crosbie.[14] However, TSB is standing strong. The migration was merely an obstacle they had to surpass on the path to their master plan. Their half-year results of 2019 indicate that they’re starting to reap the benefits of their techno-capitalistic upgrades. The load balancer automatically allocates data via Amazon’s dominant infrastructure of switches and servers. By routing data to servers with the highest availability and closest to the user, the load balancer instantly controls loan requests, transactions and network activities of TSB customers. Connections and communication accelerate, however, the human capacity to respond obstruct the pace. Speed is boosted through the exploitation of personal data. Rather than manually verifying your identity, the app does it for you to circumvent our limited responsiveness. Time-consuming application processes are over but the ease of becoming indebted accelerates. Data exploitation in combination with the pace of automation far exceeds our sentience. TSB eventually knows what you’re going to purchase before you become aware of it. ‘We understand that no-one can really predict what the future holds. So, we created a loan to help customers borrow well,’ says TSB about their fix and flex loans.[15] Conversely, predicting the future is precisely the purpose of TSB’s new platform. As Sabadell announced on their annual investors day in 2018, TSB’s growth strategy aims to ‘provide great banking to more people’ and ‘help more people borrow well’.[16] In reality, they aim to quickly increase TSB’s customer lending by 40-50% within five years from now via the swift and pre-emptive abilities of Proteo4UK.[17]


At first, the load balancer manifested itself as one of the faults but now illuminates the intentions behind the migration. As a mediator of logistics, this type of router gradually closes the space between customer activities and debt automation. However, a load balancer is but one device that fuels the bank’s desires. Various organisations and processes constitute to the gradual optimisation of the bank’s exploitative machinery. The techno-solutionism of mega-banks, the quick fixes of IT consultancies and the commercially-driven decisions of the Financial Conduct Authority will continuously repackage their exploitative processes as upgrades for the sake of enhanced user experiences. The elusive IT changes that led TSB to Proteo4UK illuminate how banks unobtrusively work their way towards the automation of the neoliberal subject. Each technological process alone would seem benign and easily overlooked. Together they uphold the consistency through which banks perpetuate and optimise debt creation, post the credit crisis. Proteus, the almighty Greek god, was known for his ability to predict the future but wanted to hide his powers from the public.[18] His ever-changing appearance allowed him to perpetually dissimulate his true intents and escape captivity. 



[1] TSB. Twitter Post. 19 April 2018, 9:36 AM.


[2] Monaghan, Angela. “TSB chief Paul Pester steps down after IT meltdown.” The Guardian. (accessed 20 August, 2019).


[3] Stiegler, Bernard. Technics and Time, 1: The Fault of Epimetheus. Trans. Richard Beardsworth and George Collins. Stanford: Stanford University Press, 1998.


[4] Virilio, Paul. The Original Accident. tr. Julie Rose. Cambridge: Polity, 2007.


[5] King, Ian. Video. “Paul Pester may have left TSB earlier but for MPs' demands he quit.” Sky News. (accessed 20 August, 2019).


[6] IBM. “IBM CIO Leadership Office - Update for TSB Board.” (accessed 20 August, 2019).


[7] Dash, Eric. "If It's Too Big to Fail, Is It Too Big to Exist?" New York Times. (accessed August 17, 2019).  


[8] Jones, Rupert. “TSB returns to the high street as Lloyds splits its branches.” The Guardian. (accessed August 17, 2019).


[9] The Economist. “The odd couple.” (accessed August 17, 2019).


[10] Ibid.

[11] Gibbs, Samuel. “Warning signs for TSB's IT meltdown were clear a year ago – insider.” The Guardian. (accessed 20 March, 2019).


[12] IBM CIO Leadership Office - Update for TSB Board, IBM, (accessed 20 August,2019).


[13] Ibid.


[14] Dunkley, Emma. “TSB’s tech system still in need of ‘400 fixes’, says new boss Debbie Crosbie.” The Times. (accessed August 17, 2019).


[15] TSB. “Personal Loans.” (accessed 20 August, 2019).


[16] Sabadell. “Sabadell 2020 – Investors Day.” PDF file. (accessed 20 August, 2019).


[17] Ibid.


[18] Encyclopaedia Britannica. “Proteus.” (accessed 20 August, 2019).