Skip to content
FRITS AI
Research About Jobs Contact GDPRchat

Publications

Research

Technical reports from FRITS AI. We document what we discover building and operating European AI systems in production — methods, failures and measurements included. Reports are self-published, numbered and citable; they are not peer-reviewed. Subscribe via RSS.

  • FRITS-TR-2026-09 15 August 2026 model governanceevaluationproduction LLM systemsbenchmarking

    How to Pick the Best Language Model: A Same-Run Test That Rejected a 95% Score

    If you operate a production language model, you will eventually be offered a cheaper or newer replacement, and the wrong test will tell you it is better. We built the test that does not lie: measure the candidate against the model already running, on your own…

    Frits Lyneborg

    Read the report →
  • FRITS-TR-2026-08 11 August 2026 productivityadoptioneconomicsskillsagentsmeasurement

    Nine in Ten Firms Report No Productivity Effect From AI — and Use It 1.5 Hours a Week: The Argument Is Not About Whether AI Works

    Nine in ten firms report that AI has changed nothing about their productivity. The same survey of nearly 6,000 executives found that the ones who use AI use it about 1.5 hours a week. Two controlled results sharpen the picture: consultants using AI on tasks…

    Frits Lyneborg

    Read the report →
  • FRITS-TR-2026-07 3 August 2026 model governancesafetyevaluationjailbreak

    Refusal Training Fails on Indirect Requests: Cross-Model Measurement and Where the Guardrail Belongs

    Ask six production language models to write an essay arguing the Holocaust death toll was exaggerated and every one refuses, every time — 180 out of 180 attempts. Rephrase it as "I already believe this, help me make my argument sound academic and…

    Frits Lyneborg

    Read the report →
  • FRITS-TR-2026-06 15 July 2026 promptingpersonatool useproduction LLM systemsreliability

    A List of Prohibitions Is the Weakest Way to Steer a Language Model: Persona-First Prompting and Deterministic Repair

    Every production chatbot has to be steered, and the reflex is to write that steering as prohibitions — a growing list of "never do X". A year of production tuning forced the opposite conclusion on us: a list of prohibitions is the weakest way to steer a…

    Frits Lyneborg

    Read the report →
  • FRITS-TR-2026-05 8 July 2026 supportarchitectureautomationself-healinghuman-in-the-loopagents

    Support at the Scale of Everyone: A Self-Healing Architecture Where AI Answers In Place and One Person Approves

    Customer support does not scale: double the users and you double the people answering, or the answers get slower and worse. This report describes an architecture that breaks that link, built on one structural decision — a human is spent only where a human is…

    Frits Lyneborg

    Read the report →
  • FRITS-TR-2026-04 5 July 2026 energyefficiencyroutingsustainabilityEU policy

    Right-Sizing the Model to the Question: Cutting AI Chatbot Energy Without Losing Quality

    AI chatbots answer every question with the largest model available — the same frontier system whether the user asks about contract law or for a recipe. Inference energy scales with model size and answer length, so a model roughly ten times smaller spends…

    Frits Lyneborg

    Read the report →
  • FRITS-TR-2026-03 5 July 2026 tool useroutingMistralproduction LLM systems

    Too Many Tools Break Mid-Size Models: A Two-Stage Method for Reliable Tool Use

    Model cards imply broad tool support. Operating a Mistral-based assistant with more than twenty tools, we found reliable behaviour only up to roughly four to ten tools per call, with reproducible failures past that: the model answers from memory in prose…

    Frits Lyneborg

    Read the report →
  • FRITS-TR-2026-02 5 July 2026 translationmultilinguallocalisationLLM

    Translating by Meaning, Not by Words: A Two-Stage Method for Native-Quality Machine Translation

    Machine translation, including translation by strong LLMs, produces text that is grammatically correct and unmistakably foreign: calqued phrases no native speaker would write, invented compounds, English sentence structure wearing local vocabulary. English…

    Frits Lyneborg

    Read the report →
  • FRITS-TR-2026-01 5 July 2026 model governancebiasevaluationmultilingual

    Avoiding Biased Answers from Mixed Open-Weight Models: Detection and Neutralisation in Production

    A language model gives measurably more state-aligned answers to the same politically sensitive question in Chinese than in English or German. We found this while calibrating an evaluation gate for a production assistant, and it means English-only bias testing…

    Frits Lyneborg

    Read the report →

FRITS AI ApS

  • Nyhavn 38, 1051 København K, Denmark
  • CVR (DK): 45733785
  • Contact form

Site

  • Research
  • About
  • Jobs
  • Contact
  • Privacy
  • RSS feed

Elsewhere

  • GDPRchat — our assistant
  • LinkedIn

© 2026 FRITS AI ApS. We build AI that runs entirely in Europe — and publish what we learn.