{"id":14179,"date":"2025-11-28T08:02:53","date_gmt":"2025-11-28T07:02:53","guid":{"rendered":"https:\/\/www.moresophy.com\/?p=14179"},"modified":"2026-03-30T13:01:32","modified_gmt":"2026-03-30T12:01:32","slug":"reliable-ai-systems","status":"publish","type":"post","link":"https:\/\/www.moresophy.com\/en\/blog\/data-driven-business\/reliable-ai-systems","title":{"rendered":"Calculating Instead of Guessing \u2013 What Reliable AI Systems Look Like"},"content":{"rendered":"\t\t
\n\t\t\t\t
\n\t\t\t\t\t
\n\t\t
\n\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\n\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t\tData-Driven Business<\/span>\n\t\t\t\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

Prof. Dr. Heiko Beier<\/a><\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t

\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

|<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t

\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\tNovember 28, 2025\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

|<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t

\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
6 min. read<\/div>\n\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

Part 2: Calculating Instead of Guessing \u2013 What Reliable AI Systems Look Like<\/h1>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\n\t\t\t\t\t\t\n\t\t\t\t\t\t\n\t\t\t\t<\/path><\/svg>\t\t\t<\/span>\n\t\t\t\t\t\t\t\t\tBack to the blog overview<\/span>\n\t\t\t\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t<\/path><\/svg>\t\t\t\t\t\t\t<\/span>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t<\/path><\/svg>\t\t\t\t\t\t\t<\/span>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t<\/path><\/svg>\t\t\t\t\t\t\t<\/span>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t<\/path><\/svg>\t\t\t\t\t\t\t<\/span>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t
\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

In the first part of this series<\/a>, we explained why the foundation for the high failure rate of AI projects is often laid in the decision-making phase for AI: Companies ask the wrong questions. They ask “What can we use AI for?” instead of “What problem do we want to solve with AI?”. <\/span><\/p>

But even if the right questions are asked in the right order, a serious problem remains – and that is the subject of this second part:<\/span><\/p>

How do you ensure that AI systems are reliable?<\/span><\/b><\/p>

This is THE critical question, especially for financial companies, but also for all other companies juggling risks and compliance requirements.<\/span><\/p>

The Reliability Problem: Why Generative AI Alone Cannot Be the Answer<\/span><\/h3>

In April 2025, BCG published an insightful <\/span>Briefing on Artificial Intelligence<\/span><\/a> and in particular on “AI Agents and the Model Context Protocol”. The central finding is worrying: the reliability of fully autonomous AI agents is rated as “low-medium”. <\/span><\/p>

And even more importantly, reliability decreases drastically with the scope of the task. After one hour of running time, we are down to 50% reliability. At over 4 hours, only 10%. <\/span><\/p>

This is not a theoretical gimmick. In concrete terms, this means that you cannot simply integrate generative AI into your core processes and hope that it will work. Especially not in an end-to-end agent system with many sequential steps. <\/span> Risk assessment in the financial sector, for example <\/span>is <\/span>always qualitative <\/span>and<\/span> quantitative<\/span>quantitative. <\/span><\/p>

Why Generative AI Is Not Reliable Per Se<\/span><\/h3>

Quantitative analyses<\/span> can <\/span>Gen<\/span>erative AI<\/span> not. <\/span>It is not reliable per se. <\/span> That sounds provocative, but it’s a fact. Why? <\/span><\/p>

Because the input in your company is something that AI has never seen before. Your data. Your organization. Your processes. Your expectations. None of this is included in the model’s training data. <\/span><\/p>

And with generative AI systems, it works like this: they deliver what sounds plausible, not what is true. They are “made to please you”. They are trained to be as helpful and satisfactory as possible – even if the facts are different. <\/span><\/p>

This is the phenomenon of “hallucinations” that everyone is now familiar with. A model invents facts. It cites literature references that it has never read. It assures you that something is so, even though the opposite is the case – with absolute conviction. <\/span><\/p>

Now you may be asking yourself: But these hallucinations are known, why aren’t they simply corrected?<\/span><\/p>

Because they are in the nature of technology. Generative AI models work according to the principle: “What is the most likely next sequence of tokens?” They do not optimize for truth, but for probability and plausibility. <\/span><\/p>

The Solution: Context Engineering and Hybrid AI<\/span><\/h3>

How do you solve this? The BCG briefing shows that an enormous methodical effort is required. But not with more generative AI – but with a well thought-out combination of analytical and generative AI. <\/span><\/p>

Here are the four pillars of reliable AI systems:<\/span><\/p>

1. Calibrated Models and Human-In-Loop<\/span><\/h6>

Models have to learn when they should trigger a human. Not all decisions can be automated. A calibrated system recognizes: “In this situation, I don’t have enough confidence to decide on my own. I’m escalating to a human.” <\/span><\/p>

2. Confidence Scores<\/span><\/h6>

Each answer from the system should have a confidence score – a numerical rating of how confident the model is with the answer. It’s not perfect, but it gives you an indication. <\/span><\/p>

3. Constraint Rules<\/span><\/h6>

Clear rules define what the agent may release and when, and when it must escalate. These are not simply technical rules – they are business rules that reflect your risk management. <\/span><\/p>

4. Analytical AI as a Foundation<\/span><\/h6>

And here comes the crucial point: all of this only works if you have a foundation of analytical AI.<\/span><\/p>

The Symbiosis: Analytical AI + Generative AI<\/span><\/h3>

At moresophy, we call this the symbiosis of analytical and generative AI.<\/span><\/p>

Analytical AI works like hard-working ants:<\/span><\/p>