Articles on the ethics and legislation of artificial intelligence, written for people who do not work in the field.
About these English versions
A weekly column on the ethics and law of artificial intelligence. These are the pieces in English — the ones on regulation, trust and safety, which travel beyond where they were written. The full archive runs to forty-odd articles in Catalan and Spanish.
An attack on Brevo altered forms on thousands of websites, WhatsApp Business now charges for some messages, and I correct two facts from the previous article.
Read the article →Three simple steps to know when to reorder before you run out: the minimum, the alert and what to watch, with the example of a hardware shop.
Read the article →Fake QR codes arrive by email, inside PDFs, or stuck over your own. The warning signs, the 30-second checks and what to do if you have already scanned one.
Read the article →AI news of the week: reporting cyberattacks within 24 hours, WhatsApp Business charging per message from October and a free guide for small businesses.
Read the article →The Catalan Government creates an AI and labour-market observatory, Spain announces a €600 million business voucher and the data-centre debate heats up.
Read the article →Scams posing as AI tools, Spain's Kit Digital grant now covering AI, and the obligation to say so when a chatbot serves your customers.
Read the article →Data centres in Òdena and Jorba, the ACCIÓ AI voucher that has already closed and the Móra la Nova gigafactory. What is real for a small business.
Read the article →Fake emails from official bodies no longer have typos or clumsy designs. The signals that still work, the two-minute checks, and what to do if you have already clicked.
Read the article →Most disputes do not start in bad faith — they start in a conversation nobody wrote down. How to keep a record, and what you must disclose before recording anything.
Read the article →This week's AI news for small businesses: an attack run by AI agents in 10 hours, how long your provider keeps your data, and introductory prices.
Read the article →This week's AI news that affects a small business: AI Act deadlines, WhatsApp charging from 1 October and three AI assistants down at once.
Read the article →Few customers leave angry: most just go quiet. How to build a list of those who have gone too long without buying, and what the automation must never do.
Read the article →An empty hour is never recovered. The three moments of a reminder that works, why cancelling has to be easy, and what the automation must never do.
Read the article →Money already earned is the worst-watched capital in many small companies. How to build a list of due dates that fills itself, and what the automation must never do.
Read the article →Delivery notes and contracts are the papers that lose the most money without anyone noticing. How to digitise them properly and what to watch.
Read the article →AI builds an impeccable fake website in an afternoon. The warning signs and the ten-minute checks to verify a supplier or a shop is real before you pay.
Read the article →A simple system with AI to publish every week without spending hours: topics from real customer questions, drafts you only have to review, and data kept under control.
Read the article →Let the new person ask an assistant instead of interrupting the team: onboarding by role, with data protected and kept inside the EU.
Read the article →A realistic guide to AI for businesses in Igualada and the Anoia region: where to start, what it really saves and which mistakes to avoid.
Read the article →How to stop a contact getting lost between messaging, email and the web form, with a weekly summary of who is waiting for a reply.
Read the article →Customers ask at ten at night. What an out-of-hours assistant should do — and what it must never do — with a practical example.
Read the article →The three real risks when a team uses AI with no rules, and why banning it does not work.
Read the article →If it is always the same person answering the team's questions, you have a knowledge-distribution problem. How to build a knowledge base with AI on your own documents.
Read the article →With ParsIA you can send invoices by WhatsApp or Telegram and they get logged by themselves. No lost receipts, no lost Saturdays.
Read the article →Whoever answers first usually gets the job. What a bot that replies to quote requests does and must not do, with a practical example.
Read the article →Bad spelling used to give a scam email away. It does not any more. What to look at now that the text proves nothing, and the rule worth more than any detector.
Read the article →Malware dressed as an AI tool, reading everything you do in the browser. The warning signs, the five-minute checks and the one rule that protects a team.
Read the article →Generative AI has made phishing fluent, personal and cheap to run at scale. The warning signs, the four measures worth the effort, and the order to act in.
Read the article →What to do in the first hour, the first four and the first day. The difference is not how sophisticated the scam was — it is the clock.
Read the article →A polished email written with AI is not a lie; an invented review is. Where the line sits when something goes out under your name, and the test that settles most cases.
Read the article →Automated moderation gets things wrong at scale. What the Digital Services Act requires, why the appeals queue is the only honest feedback loop, and what has to be reconstructable six months later.
Read the article →The Digital Omnibus postponed high risk, but not transparency: Article 50 has applied since August. The four duties, whose each one is, and what happens on 2 December.
Read the article →A practical guide after the July 2026 Digital Omnibus: what applies to you today, what moved to December 2027, and six steps to comply.
Read the article →It applies to every organisation using AI, with no minimum size. What it actually requires, what is fear marketing, and the three things to do.
Read the article →AI answers false things with complete confidence. What must always be verified and how to explain it to your team without banning anything.
Read the article →Live video deepfakes have reached business calls. The warning signs and the verification rule that stops the scam working.
Read the article →Scammers already clone voices with AI to request urgent payments. The warning signs and the simple protocol that saves you the trouble.
Read the article →How to tell somebody is using your company's name to defraud others, the ten-minute monthly check, and what to do in the first hours.
Read the article →How to spot fake reviews generated with AI, what to do if your reputation is attacked, and why buying reviews never pays off.
Read the article →Sending a photo of your ID carries real risks. Where it is justified and where it is not, and six practical habits to protect your identity.
Read the article →The five mistakes that come up again and again when a small company starts with artificial intelligence, and what to do instead.
Read the article →Keying invoices by hand is one of the tasks that steals most time from small companies. How to automate it safely, with a practical example and the privacy points to watch.
Read the article →