arm of DB. By 2028 it plans to challenge DB’s dominance along two routes: Munich-Cologne-Dortmund and Munich-Berlin-Hamburg. On July 20th it signed a €3bn ($3.5bn) deal with Siemens, an engineering conglomerate, to buy 26 Velaro high-speed trains, with an option for 14 more. “Italo’s entry is an opportunity for Germany,” says Christian Böttger of the University of Applied Sciences in Berlin. Trenitalia, Italy’s state-owned railway operator, once fiercely resisted Italo’s intrusion. But when the newcomer started running in 2012 it had a hugely beneficial effect on the country’s high-speed market. Within five years prices dropped by 40% on average, and the number of passengers more than doubled. Luca di Montezemolo, the company’s co-founder, says Germany resembles Italy before Italo. James Hanratty of Trainline, an online train-ticket platform, thinks competition is the way forward for European countries where state-owned operators still reign. Trenitalia entered France in 2021 with Frecciarossa, a high-speed service from Paris to Lyon. By December 2025 it had expanded from nine to 14 round trips on weekdays. Renfe, the state-owned Spanish operator, entered France in 2023 with its Marseille-Madrid and Lyon- Barcelona services. Average prices that year dropped by 67% on the Marseille-Madrid route and by 17% from Lyon to Barcelona, compared with 2022. In Spain high-speed rail liberalisation worked better still. In 2020 Renfe launched Avlo, a low-cost brand. In 2021 came Ouigo, the low-cost service of SNCF, France’s state-owned operator, followed a year later by Iryo, which is majority owned by Trenitalia. Between 2019 and 2024 prices fell by 35% on the Madrid-Valladolid route, 33% on the Madrid-Murcia route and 29% between Barcelona and Seville. Frequency on the three main high- speed corridors (Madrid-Barcelona, Madrid-Valencia/Alicante, and Madrid- Seville/Málaga) rose from 78 to 115 trips a day. The national competition authority estimates that improvements in rail from 2019 to 2024 attracted 4.8m passengers who would previously have travelled by road or air. Vueling, an airline, stopped flying from Barcelona to Madrid. In Sweden and the Czech Republic, says Mr Böttger, competition has triggered innovation. The Czech Republic has one of the most open markets
in Europe. Czeske Drahy, the state-owned incumbent, still has the largest market share, but it is under constant pressure from RegioJet and Leo Express, two private Czech operators. On June 25th Leo Express launched a service from Frankfurt via Prague to Przemysl, a Polish city on the border of Ukraine. Tickets for the 18-hour one-way journey start at just €10 ($11.50). The route shows how private operators can tackle one of European rail’s big problems: on a divided continent, routes that cross multiple borders are typically underserved by state-owned railways. Not all countries need competition to achieve excellence. Switzerland's state-owned operator offers some of the best service on the continent. But Switzerland is small and densely populated, and has invested in its rail network lavishly and continuously for decades. Mr Bilger has told Germans not to expect any miracles. But he might find that competition from Italians does wonders to help reform DB. And both the state-owned incumbent and Italo could soon get more competition from Italy. Trenitalia is exploring an entry into Germany’s high-speed railway market too. ■ This article was downloaded by zlibrary from https://www.economist.com/europe/2026/08/06/italians-may-help-germany-get-its- trains-to-run-on-time
The tragedy of the commons, AI edition From parking tickets to planning, the state is under siege by AI As the climate hurts Britain’s wheat farmers, it’s time to adapt The English Channel is quieter than usual. Why? Millions play padel. Will anyone watch it? If you want to understand modern Britain, listen to BBC Radio 2 Zack Polanski and the politics of eternal adolescence
The tragedy of the commons, AI edition Britain’s employment courts are clogged with AI cases Aug 6th 2026 British employment law contains a provision called “interim relief”, an emergency measure under which a judge can order a firm to reinstate a fired employee, or at least pay their wages. The subject may be a whistleblower who has complained of safety breaches, or a troublesome trade-union official. Little known outside legal circles, this provision has been sought infrequently—across Britain tribunals used to get about 20 applications a year—and rarely granted. Until recently. Data are patchy, but the surge is unmistakable. Now, around 20 applications are lodged each month in each of the 12 regional offices of Britain’s employment-tribunal system—a more than 100-fold increase— according to a memo on June 22nd by Barry Clarke and Susan Walker, the
two presidents of the system. Most of these efforts will eventually fail, but all properly filed ones are entitled to an emergency hearing and their day in court, causing delays to other cases. Judges are cautious folk but they have a prime suspect: artificial intelligence. Interim relief is a case study of how AI, like a heat-seeking missile, can lock on to the most obscure provisions of the law—and create carnage. The impact on Britain’s employment tribunals (courts that resolve disputes between employers and workers) illustrates a phenomenon emerging everywhere. AI-induced demand is overwhelming bureaucracies built for the analogue age—from Dutch municipal-tax appeals to the Canadian privacy regulator to parking-ticket tribunals in every major city. In Britain workers now ask large language models, rather than human lawyers, to help them sue their bosses quickly and cheaply. Claims have surged and backlogs grown. A case filed today may not be heard until 2030. Free, AI-powered legal advice should be good news for workers. Instead, it is proving to be a tragedy of the commons. For workers with genuine grievances, the surge in demand means longer waits for justice. For employers, it means bigger legal bills to respond to claims, both well- founded or fantastical. In the age of AI, a system intended to provide access to justice suffers from, if anything, too much access. This dynamic takes a twist in Britain, where technology and the law are moving out of step. As AI’s capacity accelerates, the Labour government is giving workers more grounds and bigger rewards for suing their employers. This means that just as demand by AI-empowered claimants is surging, the door to the court is being pushed open wider by the government.
The impact of the technology-enabled surge is striking. The number of people filing claims against their employer rose by 39% in the year to March 2026 compared with the previous year, to 50,000, while the rate of cases being resolved or rejected fell. Claims are getting more complex: the share of “open track” ones—covering issues such as sex and age discrimination— grew from 33% in 2020-21 to 61% last year. Having always taken up more court time, these types of cases now last even longer. As a result, the backlog of all unresolved individual claims rose by 55% in a single year, to 64,000 cases (see chart). AI is clearly the culprit, say lawyers, not least for the growing complexity of cases. Other explanations can be discounted. The economic downturn that usually precedes a rise in disputes is “not at all apparent”, Mr Clarke said recently. Despite unfilled vacancies, the number of judges and sitting days in court is above pre-pandemic levels. Amid all this, the quality of the process seems to have been maintained; if you are willing to wait, the judges still give a Rolls-Royce service, says one barrister. For litigants-in-person, as claimants without lawyers are known, AI can take the hard work out of a claim—and encourage them to over-egg the pudding. Enter a vague grievance into ChatGPT and before long it offers to draft a claims form, a “model legal argument” and an “employer defence prediction
map”, with tips for answering cross-examination. When our fictional claimant said they had been bullied for liking horoscopes, which is not yet a protected belief under discrimination law, the model helpfully noted that veganism is. Rhetorical flourishes were thrown in: unprompted, it wrote that our claimant had “experienced distress” and “found it difficult to obtain new employment”. The top judges are not Luddites. AI promises great efficiency gains in things like translation and scheduling hearings, Lady Carr, the head of the judiciary for England and Wales, told Parliament recently. “We are talking about AI doing the laundry so that judges can do the art.” And, her colleague Sir Geoffrey Vos has noted, AI submissions can be more coherent than the ramblings of folk “arriving at court with piles of loose papers in carrier bags”. But recent tribunal rulings reveal a growing judicial exasperation. Employment judges like submissions to be concise and factual; AI- generated claims can run to many hundreds of pages. They are often scattergun, citing scores of grounds, and packed with hallucinated laws. “We get Magna Carta, the European Convention on Human Rights and all sorts thrown at us,” says John Bowers, a barrister. AI flatters claimants who need frank advice. An NHS employee whose case included 67 grievances over 282 pages, prepared with Grok, told the judge he planned to rely on only 10% of them but did not know which. A secretary, who used ChatGPT, left the judge with the “strong feeling” that she was “pursuing a claim she does not understand and cannot personally justify when asked”. A far-right activist, also AI-assisted, presented hundreds of covert recordings of colleagues that he thought proved discrimination but which the judge found completely inaudible; he seemed “unable to grasp the hopelessness” of his case. The tech is a productivity booster for pests. One, citing discrimination due to neurodiversity, told the judge he had made over 100 tribunal claims; and yes, he had recently discovered AI. Another, subject to a restraining order by the attorney-general after filing at least 60 claims, said he would first test their merits with AI. “However,” concluded Mr Justice Griffiths, “almost all of them have been unsuccessful.”