Trump Renames AI at UN, ENISA Warns on China Hackers

Samantha: Welcome to The State of Tech, The European Edition, Wednesday September twenty-three, 2026. I'm Samantha Lawrence.

Bob: And I'm Bob Russell. Today: Donald Trump renames artificial intelligence and rejects global oversight at the UN, UK Prime Minister Andy Burnham pushes back with a call for global AI safety rules, Washington proposes an AI incident alert line with Beijing, and China-linked hackers step up cyberattacks on European shipping. Then two things you can try today: the free open-source password vault Padloc, and the offline European hiking app Cartograph. Let's get into it.

Trump renames AI to "super intelligence" and rejects international oversight at the United Nations.

Samantha: The big one comes from the UN General Assembly in New York. US President Donald Trump used a fiery speech to announce that his government will officially rename artificial intelligence to, quote, super intelligence. He argued the word artificial made the technology sound fake, and that super intelligence is a more accurate label. In the same breath he rejected calls for international rules, calling them a globalist scheme to control the technology. This came one day after more than twenty countries, including Australia, signed a joint statement pushing for greater AI oversight.

Bob: The rebrand is largely rhetorical, but the policy signal is not. Trump is drawing a clear line: no external body, no shared framework, no seat at the table for the UN on how the US develops frontier AI. That matters because the US is home to the biggest model builders on the planet, from the West Coast labs to the hyperscalers. If Washington will not sign up to a common rulebook, any global agreement immediately loses its most important signatory. For companies operating across borders, that raises the risk of a patchwork: strict rules in Brussels, looser rules in Washington, and something else again in Beijing.

Samantha: And the timing is pointed. Trump spoke on the same platform where over twenty nations had just called for more coordination. Rejecting that from the UN podium turns what could have been a diplomatic disagreement into a public confrontation. The word super intelligence also carries baggage. Inside AI research it usually refers to systems that surpass human capability across the board, a scenario that even the labs themselves say is not here yet. Adopting it as an official government term blurs that line for the public.

Bob: For European users and companies, the practical impact is indirect but real. European firms building on US models will still have to comply with the EU AI Act, regardless of what Washington calls the technology. So compliance costs and legal exposure stay in Europe, while the freedom to move fast stays in the US. European leaders have been championing greater oversight, and this speech deepens the philosophical clash. Expect Brussels to lean harder on its own framework as the anchor for anything international, precisely because the US anchor just cut the rope.

UK Prime Minister Andy Burnham calls for global AI safety rules, directly challenging Trump.

Samantha: Staying at the UN, UK Prime Minister Andy Burnham took the stage and delivered what was clearly a direct response to Trump. Burnham told the General Assembly the UK wants to lead in setting a unified set of global principles and standards for frontier AI development. He argued governments must regulate AI safety rather than leaving it to chance, and pitched the UK as the natural broker for an international agreement, pointing to its current G20 presidency as the vehicle.

Bob: The contrast could hardly be sharper. On one podium, the US president calls oversight a globalist plot. On the next, the British prime minister calls it essential. Burnham did not name Trump, but he did not need to. His framing, that safety cannot be left to chance, is precisely the argument Trump had rejected hours earlier. For the UK, this is also a positioning move. Post-Brexit, London has struggled to define its role between Washington and Brussels. Frontier AI safety, following the AI Safety Summit tradition Britain started, is one of the few files where it has genuine convening power.

Samantha: For European countries, Burnham's speech is useful. It gives them a like-minded partner outside the EU who is willing to say out loud what many capitals think but are reluctant to declare from the UN podium. It also opens a possible bridge: an EU framework, plus a UK-led global principles track, could together form the backbone of anything the rest of the world signs onto.

Bob: The open question is what Burnham can actually deliver. Global principles without the US and China are principles without teeth. Beijing has signalled openness to bilateral talks with Washington, which we will come to next. If those talks produce anything concrete, the UK-led track risks being sidelined. If they collapse, Burnham suddenly has a real opening. Either way, for European AI developers this is another sign that the regulatory landscape they operate in is going to keep shifting for at least the next year.

Samantha: Quick interruption. If you listen to The State of Tech regularly, hit that like button and subscribe, that way you'll never miss an episode. Okay, moving on.

The United States proposes an AI incident alert system with China ahead of a Trump-Xi summit.

Bob: Washington has floated a new notification mechanism for AI incidents that could affect national security. Treasury Secretary Scott Bessent confirmed the proposal, saying it was part of preparatory discussions ahead of a summit this week between President Trump and Chinese President Xi Jinping. The idea is to create a formal channel between the two AI superpowers to flag serious incidents, share information, and build a baseline of trust and cooperation around the risks of the technology.

Samantha: The parallel that comes to mind is nuclear hotlines. During the Cold War, Washington and Moscow set up direct lines precisely because the cost of a misunderstanding was catastrophic. The logic here is similar, if narrower. If a US or Chinese AI system triggers a security incident, whether that is a critical infrastructure failure, a cyber event, or something involving autonomous systems, both sides want a way to talk before it escalates. The fact that this is being proposed at all is notable, given how strained the broader tech relationship is.

Bob: There is also an obvious tension with the story we just covered. Trump rejected multilateral AI oversight at the UN, but is willing to build a bilateral channel with China. That tells you something about how Washington sees this: not as global governance, but as great-power management. The UN gets nothing, Beijing gets a phone line. For Europe, that is uncomfortable. Any US-China arrangement on AI incident reporting could quickly become the de facto international standard, simply because those two countries host most of the frontier development.

Samantha: Which means European regulators, who have spent years building the AI Act and the AI safety infrastructure around it, could find themselves adapting to a framework they had no hand in designing. Brussels will want to plug in somehow, perhaps through the AI safety institute network that already links the UK, US, Japan and the EU. But the political weight will sit with whatever Trump and Xi agree, if they agree anything at all this week.

China-linked hackers escalate cyber attacks on the European shipping sector, ENISA warns.

Bob: Now to something concrete and European. The EU's cybersecurity agency, ENISA, has issued a warning that Chinese state-linked hackers are escalating sustained cyberespionage campaigns against maritime organisations across Europe. Shipping has emerged as a consistent target. Groups such as Mustang Panda, also known as Earth Preta, have run continuous campaigns across at least seven EU member states, using spear-phishing, compromised USB drives, and customised versions of malware families including one called PlugX.

Samantha: Shipping is a soft, high-value target. European ports handle a huge share of global trade, and the sector runs on older systems knitted together with newer software. Compromising a shipping company can give attackers visibility into cargo manifests, port schedules, customer relationships, and in some cases operational technology on the vessels themselves. ENISA is essentially saying this is not a one-off, it is a sustained campaign across multiple member states, which points to strategic intent rather than opportunistic crime.

Bob: The techniques ENISA describes are telling. Spear-phishing is standard, but compromised USB drives suggest attackers are willing to go physical, targeting people who work in port terminals, on ships, or in logistics offices where thumb drives are still routinely used. That is much harder to defend against than a bad email link. Once PlugX is inside a network, it is designed to sit quietly, exfiltrate data, and stay resident for long periods.

Samantha: For European maritime companies, the practical response is unglamorous but urgent. Port operators, shipping lines, and logistics firms need to assume they are targets, not spectators. That means USB policies actually enforced, phishing training that goes beyond an annual video, and network monitoring that can spot a quiet PlugX variant. ENISA's warning also gives national regulators political cover to push harder on the NIS2 directive, which already requires critical sectors to raise their cybersecurity baseline. Maritime infrastructure is squarely in scope, and this is exactly the kind of threat those rules were written for.

The free open-source password vault Padloc keeps your logins fully in your own hands across every device.

Bob: To close out, two things you can actually use today. First, a password manager called Padloc. It is a free, open-source vault built by a small European team, and it works on Windows, macOS, Linux, iOS, Android and directly in your browser. The free tier covers a single user across all devices, which is enough for most people. If you want family sharing or a team vault, there is a paid plan, but you can start without paying anything and without giving a card.

Samantha: What sets Padloc apart from the big names is that the code is fully open-source and the company is transparent about where your encrypted vault is stored. For people who want to self-host, the server code is available too, so a technically confident user can run the whole thing on their own hardware. For everyone else, the hosted version keeps your data encrypted in a way the company itself cannot read. Reviewers highlight the clean interface, which is genuinely simpler than some of the older tools.

Bob: Setting up takes about ten minutes. Create an account with a strong master password, install the app or browser extension, and start importing from wherever your passwords live now. Most browsers and most competing managers offer a one-click export you can drop straight in. Once it is running, it will fill logins, generate strong passwords, and store two-factor codes in one place, so you stop juggling apps.

Samantha: The catch to know about: if you forget the master password, no one can recover it for you, that is the point. So write it down somewhere physical and safe before you start relying on the vault. For anyone still using the same password in two places, or a browser's built-in memory as their only line of defence, Padloc is a low-friction way to fix that this afternoon.

The offline European hiking app Cartograph gives you full trail navigation without a data connection.

Bob: Second find, and this one is for anyone planning an autumn walk. Cartograph is a mapping app built specifically for hikers, mountain bikers and cyclists in Europe. It lets you download detailed topographic maps for entire regions to your phone, so you have full navigation with no signal and no data roaming. It runs on both Android and iOS, with a free version that covers the basics and a one-off paid upgrade for advanced features. No subscription.

Samantha: What users tend to praise is the map quality. Cartograph works with several open topographic map layers, including detailed hiking and cycling maps that show contour lines, marked trails, huts, water points and viewpoints. That is a step up from general navigation apps, which know roads but not always the small paths that matter when you are on foot. You can plan a route on your laptop, transfer it to the phone, and follow it turn by turn even deep in a valley with no signal.

Bob: For European users this is genuinely practical. Wander into the Alps, the Pyrenees, the Ardennes or a Scandinavian forest, and mobile coverage becomes a lottery. Having full offline maps, plus your planned track and your current position, is the difference between a good day out and getting properly lost. The app also records your track, so you get a log of where you actually went, useful for sharing later or just remembering that great side trail.

Samantha: Practical tip: download the map tiles for your region over WiFi before you leave, not on mobile data, because the files can be large. Start with the free version, see if the workflow suits you, and only pay if you want the extras.

Samantha: Today we covered: Trump renaming AI and rejecting UN oversight, UK Prime Minister Andy Burnham pushing back with a call for global AI safety rules, Washington proposing an AI incident alert line with Beijing, China-linked hackers escalating attacks on European shipping, and two things to try yourself: the password vault Padloc and the offline hiking app Cartograph.

Bob: Want to know more or react? Visit stateoftech.eu or email us at info@doorzetters.net.

Bob: State of Tech, the tech world in 15 minutes.