The AI Act consists of rules for the use of AI in the EU. This is partly to ensure that AI systems used in the EU are safe, transparent, traceable, non-discriminatory and environmentally friendly. And to gain a better understanding of the conditions for the development and use of this innovative technology.
AI (Artificial Intelligence); Nowadays you can hardly avoid it. It’s being used increasingly and often without you realizing it, also in customer service. Usually still with good intentions, but the question is how long it will remain that way. That’s why the European Union (EU) developed a law, the AI Act (AI Regulation). What does this law entail and what will the impact be on customer service?
Purpose of the AI Act
When will the AI Act come into effect?
The AI Act was passed in the European Parliament in March 2024. Full application will follow 24 months after the date on which the regulations become active, but some parts apply earlier. Including regulations that affect the use of AI in customer service, such as meeting transparency requirements.
Voice and chatbots
In customer service, AI is used, among other things, in (voice) bots. Many companies use bots to relieve the pressure on customer service. The bots answer simple and repetitive questions, so employees are available for more complex problems. And bots also enable companies to provide customer service outside opening hours.
No more deception
As a customer you don’t always realize that you’re dealing with a bot. In recent years, developments in the field of AI have gone so fast that you can easily be fooled in terms of both image and sound. And that is where the new transparency requirements come into play. The EU wants to prevent such deception (and possible consequences thereof) and that it should be clear when you’re dealing with AI.
Notification and consent
As a company/organization, you must therefore inform customers and request permission prior to a conversation with a bot when these rules come into effect. This way they not only know that they are having a conversation with a bot, but they also give permission to enter into the conversation that way. As a result, bots will probably eventually sound and look less human to support that clarity.
Beneficial for expectation management
As contradictory as this may sound; mandatory notification and asking for permission provides benefits for people’s experience with bots. People do not expect the same results from a bot as from a human. Especially when you let the bot indicate this; “I will do my best to answer your question and if I’m not able to, I will connect you to a human colleague.” By managing expectations in this way, you increase the likelihood of a positive experience.
Options for A/B testing
This also provides opportunities for companies that have so far been skeptical about the use of bots. Mandatory notification and asking for permission is a great way to start A/B testing to discover whether your customers are open to this. Our experience shows that at least 70% of those asked choose a bot to get an answer to their question. Resulting in shorter wait times and call durations.
Also relevant outside the EU
If you’re only active outside of the EU, these developments and the AI Pact are still relevant. Chances are more countries will adopt similar measures. Even though they currently may not. And when you do, at some point, expect to provide customer service in an EU country, it is an important regulation to keep in mind.
AI is coming of age
Due to these types of developments, you notice that AI is gradually maturing. The hype is passing and it is now time to use AI in a future-proof way. So you automate where possible, but also remain personal where necessary. And you prevent misunderstandings and overly high expectations instead of feeding them. Is your bot ready for these developments?