In 2022, Starlink changed the game for maritime communications and crew welfare when it launched commercial satellite internet services for ships, using its network of now 7,800 low earth orbit (LEO) satellites.
For years, satellite communication was characterised by limited capacity, expensive and unwieldy terminals, specialised installation, multi-year contracts, sales through distributors and very high pricing. Emails were rationed, crew data was minimal, and large attachments were impossible to send or receive.
More than 60,000 ships have now installed Starlink, including most major cruise lines, and nearly all bulk carrier, tanker and offshore operators. In 2025 Starlink generated revenue of US$11 billion. The average marine subscription effectively cost around US$2,500 per month, and the hardware, a flat panel antenna, can be purchased and installed on a ship for less than US$10,000.
You can buy five Terabytes of data from Starlink for around US$5,000 per month today, more than enough for the largest crew and vessel reporting. Compared to the expensive dome antenna needed for traditional VSAT, and the lengthy fixed-term contracts for VSAT services that existed before Starlink hit the market, the company’s business model has been revolutionary. The company’s cheap, small and light hardware, high volume and high-speed data, and subscriptions, which can be flexed up or down monthly, have left Inmarsat and the incumbent satellite providers floundering.
For vessels with large numbers of passengers, like floatels, subsea vessels and survey ships, Starlink is now the essential connection that enables all aboard to browse the web, connect to their email and apps, make video calls to friends and family, and post on social media. One could also argue that nothing has revolutionised the availability of porn at sea more than Starlink, bringing home comforts to every cabin and repetitive strain wrist injury risk. Before Starlink, wifi on ships was patchy, slow and very expensive.
Starlink is owned by SpaceX, the recently listed public company with US$2 trillion of market capitalisation, which is controlled by Elon Musk, the “controversial” American businessman, Republican Party donor in the US, owner of the Tesla electric car brand, and prolific sh*t-poster (if I can use that phrase in these august pages) on the X.com website, which he also owns.
After the SpaceX IPO in June, Mr Musk holds about 46 per cent of the company’s shares, over six billion of them, making him the largest shareholder by far, and the world’s richest man. But best of all, from his perspective, Mr Musk also holds super-voting class B shares, which give him over 78 per cent of the voting rights in the company. This means he retains absolute control over major corporate decisions, and he effectively controls the Starlink signal on your ships.
In the last four years, Starlink has conquered the maritime communications world. Cue howls of anguish and angry emails from competitors trying to say this isn’t the case. Sorry, Fleet Xpress just doesn't cut it anymore.
But there is some good news. Mr Musk’s greed has created an opportunity for Starlink’s rivals to take back market share.
Starlink’s network of LEO satellites orbit the Earth altitudes of around 450 kilometres to 490 kilometres, making it cheap to launch and position them, and allowing the internet service they provide to operate with the lowest latency or lag, simply because the data takes less time to reach the satellite because it is closer to the ship’s terminal.
The LEO satellites are also much smaller than the 6,600kg satellites used for high orbit systems, with the current Starlink models weighing only 800 kg, so they cost less to launch. Their proximity allows them to complete an orbit roughly every 90 minutes, which enables rapid data transmission and minimal signal delay compared to higher-orbit satellites.
This low-latency performance is crucial for applications like real-time internet, web camming, video calls, and autonomous navigation. Anyone who has tried to make a Teams video call or hold a Zoom meeting with a ship over VSAT will appreciate the speed advantage LEO systems have over the traditional geostationary (GEO) satellites, which have typically been owned by Inmarsat for maritime communications.
GEO satellites are positioned in a circular orbit 35,786 kilometres above the equator, and they revolve in sync with Earth's 24-hour rotation. From the ground, the satellite appears stationary in the sky, allowing antennas to remain fixed without tracking its movement. Instead, with LEO systems like Starlink and OneWeb, the signal jumps from satellite to satellite as they orbit away from the ship.
That’s not to say that there aren’t flaws to Starlink, however.
Firstly, several important countries have forbidden Starlink to operate within 12 miles of their coastline in their national waters. Most important of these is China, which has fined foreign ships that do not switch off their Starlink terminals before arrival. Hong Kong and Macau have also banned it, as both fall under PRC sovereign telecommunications regulation.
Other countries that have made the service illegal, include the usual repressive tyrannies of Iran, Belarus, Russia, North Korea and Afghanistan. The Iranian Government has criminalised possession of Starlink terminals and enforces the ban with prison sentences of up to 10 years, with reports of execution penalties also being threatened. The government in Tehran deployed nationwide GPS jamming concurrent with the January 2026 internet blackout during its deadly crackdown on the student protests, which cut Starlink throughput by up to 80 per cent.
For autocracies, uncontrolled satellite internet carries the risk of subversion and rebellion. In other states, bureaucratic inertia or corrupt vested interests have also slowed down adoption.
In each country, the national telecom regulator controls three gates that any satellite broadband operator like Starlink must pass. John Maximus highlighted that these are spectrum licensing for gateway earth stations, equipment type approval for user terminals, and service authorisation to sell commercially.
Unfortunately, Starlink is stuck in regulatory limbo awaiting approval to activate the service in national waters in over thirty nations. These include South Africa, Egypt, Iraq, Lebanon, Morocco, Algeria, Tunisia, Turkey, and Taiwan. Starlink said, for these countries, the, "service date is unknown," or the service is, "pending regulatory approval." This handy map shows the status.
Liberal democracies have generally welcomed Starlink, while crooked and failed states have failed to permit its use.
In South Africa, Starlink has been blocked since launch by the state telecom regulator ICASA’s electronic communications licensing requirement that all licensees must be 30 per cent owned by effectively Black South Africans (a requirement known as historically-disadvantaged-ownership even though apartheid ended in 1994). This is part of the ruling African National Congress’ (ANC) so-called Broad-Based Black Economic Empowerment framework. Because Starlink is owned by SpaceX, it cannot meet that threshold without restructuring and selling a chunk of its local business to an ANC crony local black businessperson. As a former South African citizen, Musk has indicated he is not willing to do that.
There are also geographic limitations to Starlink’s service. Basically, in the far north and the far south of the globe, there is no Starlink signal. You can’t get it in the Arctic above 72°N, nor in the Southern Ocean and Antarctica below 56°S. And there is no gradual degradation of service; beyond those limits, Starlink just stops working.
Critically, whilst VSAT offers a committed information rate (CIR) in its contracts, Starlink's service level agreement contract does not turn the speeds advertised by the company into any contractually binding obligation. Maybe your ship will get the download speed and data volume you hope for; if not, you have no recourse against Starlink or SpaceX. A CIR written into a VSAT contract is the only way to buy a guaranteed satellite connectivity rate for a ship, which might be important if you need it for chart updates, transmitting survey data or facing a crew health issue that needs remote telemedicine.
Finally, Starlink Maritime is not type-approved for GMDSS and does not satisfy SOLAS distress and safety communications requirements, so on a commercial vessel, it must remain supplementary to approved SOLAS certified equipment, typically from Inmarsat and its GEO satellites. Where owners genuinely want to move away from Inmarsat, the only other certified alternative is Iridium’s GMDSS service, which was recognised by the IMO in 2020, rather than Starlink.
Now, Starlink’s dominance is under threat from two self-inflicted decisions.
For a start, Starlink is unilaterally raising its pricing by 50 per cent for marine customers from December onwards. Cheap fast data will remain cheap compared to Inmarsat, but it will be 50 per cent more expensive compared to today.
Ouch.
Now that SpaceX is a public company and has captured huge market share in the commercial marine satellite communications market, it intends to ramp up its marine revenues, clearly.
Some of this is driven by a change in Starlink's business model. Two weeks ago, Starlink agreed to buy a portfolio of “low-band” spectrum that would be suitable for broad mobile phone coverage in the US. Starlink spent US$8 billion on a bundle of spectrum, but in order to launch a terrestrial service, it would need to build mobile towers and antennas. To match American mobile/cellphone service coverage of companies like AT&T, Verizon and T-Mobile could cost more than US$110 billion in capital expenditure and would take around ten years, analysts have estimated. Someone has got to pay for that, captain.
There is a window for these higher prices for Starlink’s marine customers to succeed. Starlink has two new LEO competitors, but both of them have been tardy deploying their satellite constellations. Starlink’s price rise enables them potentially to come in and undercut it.
Amazon (yes, the internet retail phenomenon) received Federal Communications Commission authority to deploy 3,236 satellites into LEO in July 2020, and then confusingly renamed its Starlink rival Amazon Leo. Service is expected to begin after the first 578 satellites reach orbit, but as of July this year, Amazon Leo had deployed only 396 satellites.
Under the original terms of its license, Amazon was supposed to have launched and put half of the constellation into operation by July 30, 2026, but (surprise!) the FCC waived the July deadline in early June this year.
So, Amazon Leo is not competing yet. Its package of satellites has not left the distribution centre. Please wait for service delivery.
The second competitor is Eutelsat OneWeb, a company that has been better at sponsoring maritime conferences and delivering Powerpoint presentations than actually getting its LEO constellation into service on ships.
OneWeb was founded under the name WorldVu in 2012, then renamed itself OneWeb and launched its first six satellites in February 2019. It satellites orbit at 1,200 kilometres of altitude.
Unfortunately, OneWeb went bust in March 2020 after failing to raise the required funds needs to build and launch into space the remaining 90 per cent of the satellites needed for the LEO constellation. Readers with long memories will recall that Iridium, the first private satellite phone company using LEO satellites, met a similar fate in 1999 due to massive debt, high operational costs and limited subscriber adoption.
However, OneWeb was rescued from bankruptcy and the UK Government and India’s Bharti Global both chipped in US$500 million apiece to keep the company afloat. It merged with Eutelsat in 2023 and launched its commercial marine LEO proposition in the same year. Its parent Eutelsat initially specialised in satellite TV and radio in Europe, and is listed on the Euronext Paris stock exchange.
As of this month, 651 OneWeb satellites are in orbit so the company can offer a global footprint for offshore vessels and other ships. But launching second meant that many fleet managers had already committed to Starlink roll-outs, and did not want to the hassle of their fleets being split between two different types of hardware and two different service providers. They could see Starlink worked and was cheaper, so OneWeb remained very much an unloved secondary service.
However, Starlink’s price hikes offer OneWeb a huge opportunity to fight back, if its management can seize it and roll out its antenna on thousands of ships.
Ukrainians also discovered a horrific flaw in Starlink’s business model last week.
Then “somehow,” Starlink terminals used by Ukrainian military drone units began experiencing signal failures at speeds above 160 kph, disrupting several missions early on Wednesday (14th) morning (local time), according to both Ukrainian military personnel and senior officials. That night, Russian ballistic missiles targeted all three of Kyiv’s municipal combined heat and power plants and other energy infrastructure in surrounding regions. Attacks across Ukraine also killed at least five civilians and injured dozens, as per Kyiv Post.
The service limitations came with “no warning, just in the middle of the night”, one official told the Financial Times. Ukrainian defence ministry adviser Serhii Beskrestnov was reported by the newspaper to have said that long-range drone operations were also affected, as was Ukraine’s ability to defend itself against Russia’s new jet-powered drones, since Starlink’s Starshield service is used on the interceptors being developed to combat them.
It was then reported on Friday in the Daily Telegraph that Elon Musk had personally conspired with President Donald Trump to reduce the Starlink service in Ukraine.
Musk initially directed the blame at the Trump administration, saying that, "Starshield systems are owned and controlled," by the federal government and that the military and, "any limits on speed or usage are decided by the US military and ultimately the President."
An unnamed source told The Telegraph that after the "night of carnage" in Kyiv, Musk was reportedly scared that he would face a backlash from consumers in Europe and so the Starshield service was eventually restored to full functionality in Ukraine by Friday night.
The incident was the best advertising OneWeb could hope for. So long as Mr Musk remains involved with SpaceX, the Ukrainian shutdown showed that its products cannot be trusted to work when most needed. Today it is Ukraine, but what happens if the US has a conflict with another state? Could Starlink services be assured on vessels operated by Canadian or Mexican owners if Washington decided to ratchet up pressure on those countries?
As for European alternatives, OneWeb starts to look very attractive as a more neutral option. It seems increasingly certain that European governments have to invest heavily in alternatives to Starlink as a strategic necessity.
Starlink may have been the first mover, but both the 50 per cent price increase and the deadly decision to slow down the Ukrainian service show that it is no longer trustworthy as a service provider. The tight connection between Starlink and the American government was reinforced this weekend.
Yesterday, the Financial Times reported that the US Ambassador to the EU has warned of “devastating retaliation” if the EU went ahead with plans to limit spectrum access for non-EU companies like SpaceX. The European Commission had proposed earlier this year that part of its spectrum band that allows smartphones to connect directly to satellites for European players should be restricted to EU companies only. That would limit the frequencies for both Chinese and US companies.
The ambassador’s intervention was another reminder that Starlink is not just a convenient service for seafarers to surf the web. It is also a potential tool of leverage by the US Government over foreigners, all seven billion of them.
This month, it was the Ukrainians who lost out to interference by Starlink and Washington. Many will be asking who is next and what the alternatives are. This gives Starlink’s competitors an opportunity. Let’s see if they can take it.
Meanwhile, brace yourself for smaller internet quotas and pricier top-ups onboard from December onwards, Mr Musk’s happy Christmas to seafarers everywhere.
Further reading
An analysis of the 1999 bankruptcy of Iridium is here. Salutary for those other operators who look at Starlink and see an opportunity. This is a capital-intensive business with very high fixed costs. You either have a functional LEO constellation, or you don’t; launching a few satellites and then running out of money is disastrous, as is launching all your satellites and then not being able to service your debt because you don’t have the subscribers.
Compared to the cost of installing and providing airline wifi, even with the coming price hike, Starlink remains cheap for maritime users. A business jet can pay US$20,000 a month for Starlink connectivity. Reuters cited a Jeffries report that estimated that American Airlines' planned rollout of the service on 500 of its narrowbody planes could cost US$150 to US$250 million to install, plus more than US$60 million every year in service fees. This month, American Airlines said it was going full-on Starlink and would increase the roll-out to one thousand of its planes. Starlink already has contracts covering more than 7,000 aircraft.
Signal Before Noise has an excellent article on why OneWeb has not achieved remarkable success, which sets out that OneWeb’s management is still wedded to a business model like that of Inmarsat, whereas Starlink has revolutionised the mindset of users towards contract flexibility, cheap and easy hardware and simple pricing models.