AI-policy för företag: mall och guide
Anställda använder redan AI privat. En tydlig AI-policy flyttar vanan in i arbetet – utan att bryta mot GDPR. Så skriver du en, med mall och exempel.
Reference material on Claude and AI adoption, kept current — plus what changed along the way.
Reference material we keep current — the things worth coming back to.
Anställda använder redan AI privat. En tydlig AI-policy flyttar vanan in i arbetet – utan att bryta mot GDPR. Så skriver du en, med mall och exempel.
What Claude's Research mode does, where to find it, when it earns its keep, and when a normal chat is the better tool.
Does Claude train on your data? On business accounts, no — contractually. On personal accounts it differs. The distinction most teams never check.
What Claude connectors are, how to set one up in two minutes, which tool to connect first, and what MCP actually means in plain language.
What Claude's Extended Thinking does, why it differs from picking a bigger model, when it changes the answer, and when to leave it off.
Claude Reflect shows you how your team actually uses AI. What it tracks, what it reveals about your habits, and how to read it — in plain English.
What Claude Cowork does, how it differs from Claude Code, where it runs, and the first three tasks worth handing to it.
Claude Science is Anthropic's workbench for research teams. What it does, who it's built for, and why it matters even if you're not a scientist.
What the Claude Chrome extension does, how it differs from Computer Use, which browser tasks to hand it first, and how to use it safely.
Sonnet by default, Opus when being wrong is expensive. A task-by-task table for choosing between Claude Opus, Sonnet, Haiku and Fable.
Anthropic's models are coming to Apple devices. What the Foundation Models integration changes for everyday work, and what to expect when.
What Claude's context window is, what a million words actually holds, which tasks it unlocks, and why more context is not automatically better.
Anthropic just shipped self-hosted sandboxes and MCP tunnels for Claude Managed Agents — the features that answer the question every cautious business…
Anthropic trains Claude on a written constitution. What's actually in it, how it shapes every answer you get, and what that means for your team.
What Claude Artifacts are, what Claude can build with them, how link sharing works without an account, and how they differ from Projects and Skills.
What Claude Skills are, the test for Skill versus Project, and how to build your first one backwards from work you have already done well.
What Claude Projects are, how to set one up, what to write in the instructions field, and how Projects differ from Skills and memory.
What Claude’s memory stores, how to turn it on, read it, correct it or delete it — and the difference between personal and agent memory.
A non-technical founder on learning to ship software with Claude Code — the order it actually happened in, and what to build first.
Legal AI adoption is accelerating faster than any other professional services sector. Here's where AI delivers the most value for law firms, and how to…
AI adoption in accounting firms jumped from 9% to 41% in a single year. The firms investing now are building a structural advantage. Here's where AI…
Real estate runs on relationships — but most agents spend half their time on work that isn't building them. Here's where AI saves the most time in real…
Professional services firms lose 15–20 hours per person per week to work AI can compress. Here's exactly where, how to recover it, and why the firms that…
Varför tillgång till AI-verktyg sällan blir användning, var effekten är störst i skånskt näringsliv, och hur en session går till.
The organizations seeing the biggest results from AI aren't the most technical ones. Here's what they have in common — and what it means for how you…
How SaaS companies use AI across product, customer success, sales, and operations — and why the internal adoption challenge is bigger than the product…
Deployed är baserade i Göteborg och hjälper västerhavsföretag att gå AI-native — med workshops som faktiskt förändrar hur teamet arbetar, inte bara vad de…
Letar du efter en AI-konsult i Stockholm? Deployed hjälper stockholmsföretag med 20–200 anställda att gå från AI-nyfiken till AI-native — med hands-on…
The CEO's role in AI adoption isn't to mandate tools — it's to model behavior. Here's what effective AI leadership actually looks like, and why it matters…
AI consultant pricing is confusing and rarely transparent. Here's an honest breakdown of what different AI consulting models cost, what they deliver, and…
How marketing teams use AI to produce more content, run better campaigns, and cut reporting time — without adding headcount or compromising quality.
Not all AI workshops are equal. Here's exactly what a good AI workshop looks like, what to ask before you book, and how to know if it will actually change…
The AI consulting industry is producing a lot of strategies, slide decks, and frameworks. Very few deployed tools. Here's what's broken — and what…
De flesta AI-workshops informerar men ändrar inget. Här är vad som skiljer dem som gör det, och hur en session ser ut i praktiken.
Looking for an AI consultant in the UK? Deployed runs hands-on sessions for British companies with 20–200 employees where every participant builds…
How recruitment agencies use AI to source faster, screen more consistently, and place more candidates — without adding headcount or compromising on…
Looking for an AI consultant in Sweden? Deployed helps Nordic companies go from AI-curious to AI-native — with hands-on workshops and ongoing support that…
Why employees resist AI, what makes it worse, and the order in which scepticism actually turns into use.
Claude Cowork är inte en chatbot. Det är ett verktyg som faktiskt gör jobbet åt dig — läser dina filer, gör en plan och levererar färdiga resultat medan…
Claude Code lets non-technical people build working software tools — without writing code. Here's what it is, what it can do, and why it matters for your…
How recruiters use AI to cut sourcing, screening, and communication time in half — without losing the human judgment that makes great hiring possible.
Five AI automations your sales team can build in under an hour — no technical skills required. Real examples, real time savings, ready to use tomorrow.
Where to start, the steps that matter, the mistakes that kill adoption, and how to measure whether any of it is working.
Four failure modes account for nearly every AI workshop that changes nothing — and the fix for each is the checklist to buy against.
AI-native means every person uses AI to do their actual job — not a technology project. What it looks like at 50 people, and what the path really is.
Everyone's talking about AI agents. Almost nobody's explaining why most of them break the moment they leave the demo. Here's what actually determines…
Original findings we published — data nobody else has framed.
What changed in the Claude ecosystem, and what it meant at the time.
Claude now has its own browser inside Cowork. What it does, how it differs from the Chrome extension, how logins work, and what to hand it first.
Most people tried talking to Claude once, found the answers noticeably thinner than the ones they got by typing, and quietly went back to the keyboard.…
On July 6, 2026, Anthropic published research showing that Claude has something nobody built into it: a small, private mental workspace where it holds…
Anthropic just flipped an important switch on the Microsoft 365 connector — the link that lets Claude reach into your company's Outlook, Calendar…
Anthropic and Microsoft just made Claude generally available inside Microsoft Foundry — the AI platform that lives inside Azure, the Microsoft cloud most…
Once you roll Claude out to a whole team instead of a few enthusiasts, a new question shows up: who's actually using this, on what, and what is it costing…
Anthropic just launched Claude Tag — the first version of Claude built to be multiplayer. It lives inside Slack, and anyone in a channel can type @Claude…
Anthropic just shipped scheduled deployments and vault credentials for Claude Managed Agents. Your AI agents can now run on a schedule without anyone…
On June 9, 2026, Anthropic released Claude Fable 5 — its most powerful AI model ever. Three days later, the U.S. government ordered it shut down under an…
On June 1, 2026, Anthropic filed to go public at a valuation approaching $1 trillion, with $47 billion in annualized revenue and its first profitable…
Anthropic launched Claude Security in public beta — an AI that scans your code for vulnerabilities and suggests fixes. Meanwhile, Project Glasswing has…
PwC and Anthropic announced the largest Claude training program in the world — 30,000 professionals certified on Claude, with results like insurance…
On May 12, 2026, Anthropic launched Claude for Legal — 12 practice-area plugins, 20+ MCP connectors to tools like Westlaw, iManage, and Everlaw, and 80+…
On May 13, 2026, Anthropic launched Claude for Small Business — a package of connectors and ready-to-run workflows that plug Claude directly into…
On May 6, 2026, Anthropic launched Dreaming and Outcomes for Claude Managed Agents. Your AI agents can now review their own past work, learn from…
On May 7, 2026, Anthropic launched multiagent orchestration for Claude Managed Agents. Claude can now act as a manager that delegates work to specialist…
On May 5, 2026, Anthropic launched ten AI agent templates for financial services and made Claude add-ins for Microsoft 365 generally available. Claude now…
On April 28, 2026, Anthropic launched nine new Claude connectors for creative software — Adobe, Blender, Autodesk Fusion, Ableton, Splice, Affinity by…
On April 17, Anthropic launched Claude Design — a new product inside Claude that turns prompts into finished pitch decks, landing pages, interactive…
Anthropic just released Claude Opus 4.7 — a major upgrade to the AI model that powers Claude. Better instruction following, 3x sharper document reading…
Anthropic just shipped Claude Code Routines, a redesigned desktop app, and Computer Use in Cowork. Here's what those three updates mean in plain language…
While you\'re evaluating AI vendors, your competitors are shipping. Here\'s what inaction actually costs.
Most teams never measure whether a workshop worked. The metrics that matter, how to collect them, and how to build the internal business case.
Most companies are using AI. Almost none are seeing meaningful results. Here's the specific difference between organizations that get real value from AI…