Kevin Liebholz

Transitioning to team leadership to lead, inspire, and coach teams toward excellence and sustainable, customer-driven solutions.

7 minute read

Today, I want to share my experience attending “The EU AI Act” event organized by AI Hub Frankfurt on July 18th, 2023. It was an insightful gathering that shed light on the upcoming regulatory landscape surrounding AI and its impact on businesses. In this post, I will provide you with a summary of the keynotes and discussions from the event.

This article was written with ChatGPT :robot: by providing my notes from the event and adapting the AIs answer. See my original notes and my ChatGPT prompt below the article.

TL;DR

  • Attended “The EU AI Act” event organized by AI Hub Frankfurt.
  • the EU AI Act is not through yet. But when adopted this year, there will be a grace period until 2025 (probably)
  • EU AI Act aims to mitigate AI risks and imposes penalties for non-compliance.
  • Act categorizes AI applications based on risk levels: high, moderate, minimal/no.
  • Compliance with AI standards helps control risks, build trust, and scale AI adoption.
  • Event emphasized the importance of understanding and managing AI risks.
  • Businesses should prioritize AI quality and compliance for successful transformation.
  • Steps recommended include training, assessments, advisory services, and governance.
  • Panel discussion covered trustworthiness, global enforcement, and ethical considerations.
  • Balancing regulation and practicality is key in implementing the EU AI Act.
  • Businesses must navigate regulations, prioritize ethics, and ensure data privacy.

The EU AI Act

AI Hub Frankfurt: Supporting Innovation and Collaboration

AI Hub Frankfurt serves as a vibrant community hub, fostering innovation, and providing support to startups in the AI space. With a diverse range of offerings, including mentoring, prototyping assistance, consulting, and upskilling opportunities, AI Hub Frankfurt is an excellent resource for individuals and organizations seeking to explore the potential of artificial intelligence.

One exciting announcement during the event was the upcoming AI week, scheduled to take place from October 16th to October 20th, 2023. This week-long event will feature various activities, including AI-related workshops, networking sessions, and even a hackathon. If you’re interested in participating, make sure to get in touch with AI Hub Frankfurt to secure your spot.

Keynote: Overview of the EU AI Act

The keynote presentation focused on the “EU AI Act” and was delivered by Tüv Süd. The talk highlighted the importance of understanding AI risks and pitfalls and how the EU AI Act aims to mitigate them for business success.

AI Risks & Pitfalls

The speaker emphasized the various risks associated with AI implementation, including safety, security, legal concerns, performance issues, ethical considerations, and sustainability. It is crucial for businesses to address these risks proactively to ensure the responsible and effective deployment of AI systems.

EU AI Act & Standards

The EU AI Act primarily aims to mitigate risks associated with AI implementation. It introduces a penalty of 6% of a company’s total worldwide revenue for non-compliance. However, businesses will have a grace period until 2025 if the act is adopted in 2023. Similar to the General Data Protection Regulation (GDPR), the EU AI Act applies to all companies and does not require adaptation by national governments.

The act classifies the application of AI models based on risk levels, categorizing them as high risk, moderate risk, or minimal/no risk. This risk-based approach allows for targeted regulation and ensures that higher-risk applications receive appropriate scrutiny and oversight.

AI Quality and Compliance

To achieve AI quality and compliance, businesses need to adhere to various standards, such as those defined by the International Organization for Standardization (ISO). Complying with these standards brings several benefits, including the ability to control and manage risks effectively, build trust with customers, and scale the adoption of AI technologies.

During the keynote, actionable steps were suggested for businesses to take immediately, including investing in AI training, conducting assessments, and seeking advisory services to ensure compliance with standards and regulations.

Keynote: Impact of the EU AI Act on Transformation

The second keynote presentation delved into the broader impact of the EU AI Act on business transformation. It was emphasized that the act’s mandatory requirements apply not only to high-risk AI applications but also to the entire value chain of AI systems.

Furthermore, the keynotes highlighted the close relationship between AI quality and compliance, reinforcing the idea that businesses must prioritize both aspects to ensure successful AI adoption.

To prepare for the implementation of the EU AI Act, the following steps were recommended: conducting assessments and benchmarking, establishing governance and setting directions, and investing in the development of organizational maturity in handling AI technologies.

Panel Discussion: Digital Ethics & Regulation

The event concluded with a panel discussion focused on digital ethics and regulation. The participants explored various aspects of the EU AI Act and its implications. Some key points discussed were:

  • The EU AI Act’s management of AI components in software and its impact on EU citizens’ fundamental rights.
  • The importance of trustworthiness in AI systems and how to determine which AI applications are deemed trustworthy.
  • The competitive advantage provided by trust and privacy.
  • The importance of auditability and the need for transparency and accountability. E.g.: contracts involving Chinese businesses could have a section guaranteeing that they can be audited.
  • The global enforcement of AI standards and the promotion of ethical practices worldwide.
  • The recognition that even if a business is regulatory compliant, it may still engage in unethical practices.
  • The heterogeneity of privacy regulations within the EU and the need for businesses to take responsibility for their data practices.
  • The closing question: Is the EU AI Act perceived as over-regulation or is it seen as a necessary and mandatory step for responsible AI adoption? The discussion emphasized the importance of balancing practicability and feasibility while adapting to evolving AI landscapes.

Overall, the event provided valuable insights into the EU AI Act and its implications for businesses. It underscored the significance of understanding and managing AI risks, complying with relevant standards, and fostering ethical practices to navigate the regulatory landscape successfully.

Stay tuned for more blog posts on AI-related topics and industry events. If you have any questions or want to share your own experiences, feel free to leave a comment below!

ChatGPT prompt

:robot:

imagine you are a developer who owns a more or less technical blog in order to share your expertise and experiences. One goal is to share events that you visited. You just visited the “The EU AI Act” event from the “AI Hub Frankfurt” and you took notes using the markdown language. I will provide you with those notes. Write a short blogpost from those notes using markdown.

Here are the notes:

AI Hub

  • support startups (mentoring, prototyping)
  • topics
    • AI Events
    • AI Startups & Innovation
    • AI Consulting & Support
    • AI Talents & Upskilling
  • AI week: 16.10.23 - 20.10.23 (incl. Hackathon)
    • get in touch to take part in it

Überblick EU AI Act (keynote)

  • AI quality - navigating AI regulations for business success
  • speaker: Tüv Süd

AI risks & pitfalls

  • risks
    • safety
    • security
    • legal
    • performance
    • ethics
    • sustainability

EU Ai act & Standards

  • EU Act is only about mitigating risks
  • penalty: 6% of total worldwide revenue
  • if adopted 2023 -> grace period until 2025
  • all companies (like GDPR)
  • not need to be adapted by national governments
  • classify application of the model depending on risk (high risk, moderate, minimal/no)

AI Quality and Compliance

  • lots of standards to be applied for compliance (ISO)
  • benefits
    • control & manage risk
    • build trust with customers
    • scale the adoption of AI
  • What can be done now
    • training
    • assessment
    • advisory

Impact of the EU AI Act on Transformation (keynote)

  • mandatory requirements
  • not only for high-risk but entire value chain
  • AI quality & AI compliance go hand in hand
  • What can be done now
    • assessment & benchmark
    • Set directions & establish governance
    • maturity development

Panel Discussion - digital ethics & regulation

  • EU AI Act
    • manages AI components in software
    • where are EU citizen (fundamental) rights restricted?
    • which AI is trustworthy
    • risk based approach
  • competition advantage: trust, privacy
  • contracts with Chinese bizs: add how much they can be audited
  • how to durchsetzen standards worldwide
  • even if regulartoy compliant, could be unethical (e.g. HR stuff)
  • even EU is heterogenous regarding privacy
  • cannot outsource responsibility (esp. for data)
  • Closing: EU AI Act over-regulation or mandatory?
    • but adaptable in practicability and feasibility

please add a summary to beginning of the article and provide me the article in markdown code

provide an summary including the main learnings from the article in keywords

provide an summary including the main learnings from the article using bullet points

Provide a tl/dr of the article in bullet points

Categories:

Updated:

You may also enjoy

Cheatsheet “Deploy Empathy”

7 minute read

Hello everyone, I took some notes during reading “Deploy Empathy” by Michele Hansen. If you want to learn how to talk to customers to make great business dec...