Menu
Menu
It’s nothing new that artificial intelligence (AI) is not only taking the world by storm but is also fundamentally changing it. At Deutsche Telekom, we are already using this technology in multiple ways. For example, to solve business problems or increase our productivity. This is the only way we can make things faster – and better – for our customers. However, none of this should be at the expense of the environment. That’s why we now have the “Green AI Principles”: Nine Principles for green artificial intelligence.”
Chatbots such as “Ask Magenta”, support with fiber optic expansion, or important help with cyber security – at Deutsche Telekom, we work on pattern recognition, automation and machine learning and use artificial intelligence for the benefit of our customers. We are pursuing a healthy mix of “taking”, “shaping” and “making”. This means that we purchase ready-made AI solutions (“taking”), develop them further for our needs (“shaping”), or even develop them from scratch (“making”). Of course, we want to play a leading role in the use and provision of AI for the benefit of society, the economy, and the company.
To put it in a nutshell: AI simplifies the digital transformation, supports employees and makes their work more efficient and better. For our customers, the use of AI simply makes many things faster – the expansion of fiber optics, for example. Without AI, we would never be able to meet the current level of demand at the speed that is now possible. When it comes to the important topic of cyber security, we use AI to recognize patterns in attempted attacks and ward them off. It’s all pretty smart.
But what about the environment?
Brave new AI world, but as usual, there are pros and cons here, too. Even though AI can contribute to climate protection, it consumes an enormous amount of energy and resources. So how does this work together? After agreeing on ethical guidelines for the use of AI at Deutsche Telekom back in 2018, we are now committed to a more ecologically sustainable use of the technology. We not only want to make the development and use of AI more sustainable – we also want to provide an impetus for companies and institutions, for politics and science to calibrate AI for sustainability right from the start. This is why we have developed our GREEN AI PRINCIPLES.
Good in principle, indispensable in practice
We established nine principles for green(er) AI. They are intended to show how AI solutions can be developed and used in a more environmentally sustainable way. More importantly, they show how risks – such as a significantly increasing carbon footprint – can be addressed preventively to make the technology greener from the get-go.
Find out more about the nine principles and how each of them helps to shape AI in a way that benefits people and the environment:
AI applications need electricity, that’s obvious. However, as they consume more electricity than conventional IT applications, it is even more important that they are powered by electricity from renewable energies and that harmful emissions are avoided. Incidentally, we have been following this principle in practice since 2021 – not just for the electricity requirements of AI, but for AI as a whole. And because we believe this is the way to go, we recommend that our partners do the same.
We use hardware, software, AI models, and data multiple times, which makes us more flexible and efficient. In doing so, we avoid unnecessary energy consumption.
Full transparency for the invisible danger. In our AI development, we not only monitor the carbon emissions of hardware and software – we also analyze how hardware and software changes affect our ecological footprint.
Size matters! Simple math: huge hardware also consumes huge amounts of energy. That’s why we adapt our IT equipment flexibly to our needs and dispense with everything we don’t need for modeling, training, or operation.
We prefer to use optimized and tested AI architecture and models – suitable for the application. Ideally, we can use software components several times according to the modular principle. For more efficiency and less energy consumption.
We don’t reinvent the wheel every time we develop software for new AI applications. That’s why we use synergies for similar use cases and thus avoid unnecessary duplication of work. We tell it like it is: we reuse code whenever possible.
Did you know that different programming languages have different energy consumption? That’s why we are also energy-conscious in development and therefore program as efficiently as possible.
We keep it simple: if specific tasks can be fulfilled, we tend to use the simple AI models. Within the selected models, we pay attention to efficient algorithms. By using more energy-intensive AI applications only where they are really needed, we can conserve valuable resources. Simple as that.
Last but not least, we pay attention to what happens in our value chain. Regularly checking the carbon footprint of the hardware and software of our AI applications is therefore a matter of honor. We stand by our end-to-end responsibility for Green AI with one focus on Gen AI.
Artificial intelligence leads to greater energy efficiency
For us, sustainability and AI can go hand in hand. Sustainability therefore also plays a decisive role in data centers where artificial intelligence is used to optimize their own operating processes – more energy-efficiently and with lower costs. We explain this with the example of cooling system control. An AI calculates the expected IT load and intelligently controls the cooling systems on this basis, for example by adjusting the well water cooling in the data centers.
At Telekom, we use and promote various AI solutions in the spirit of sustainability. In our mobile network, an AI solution puts antennas into “sleep mode” when they are not needed. This automatically saves a lot of unnecessary energy. The so-called “AI Vision Suite” is an example of the use of AI in process optimization. It identifies objects and analyzes image, video and 3D data in real time. This is how efficient quality control works today, as defects can be detected at an early stage causing material waste to be avoided.
Efficient, social, and green together
The topic of artificial intelligence has potential for many more articles. Sustainability is also particularly important to us here. As a key enabler of AI – thanks to our stable network – we bear a great deal of responsibility in more than one area. As always, we want to set a good example and think a little further ahead.
If you want to find out more about AI at Deutsche Telekom, you can do so here.
You are currently viewing a placeholder content from Default. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More Information You are currently viewing a placeholder content from Default. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More Information
You are currently viewing a placeholder content from Google Maps. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More Information
