# What the EU AI Act means for you—and how to prepare

## 4 key takeaways

The EU AI Act is considered the world’s first comprehensive AI
law, driving new levels of oversight and regulation. The law aims
to build trust and ensure protection among users throughout this
period of innovation and discovery. What does it mean for you
and your business?

## First, a quick primer. The EU AI Act:

– Applies to businesses operating within the EU,

including those based outside the EU if their AI,
or the outputs of the AI, are used in the EU

– Regulates AI systems based on risk level.

Some of the most important provisions include - a prohibition
on certain AI practices, standards for developing and
deploying high-risk AI systems, transparency obligations for
certain types of AI, and rules for general-purpose AI models

– Defines obligations for multiple operators in the AI value

chain, including for creaters (providers) and for users
(deployers) of AI systems

– Provisions will phase in over a period of 6–36 months,

depending on the provision

– Outlines fines as follows:

-  Up to 7% of global annual turnover or EUR 35 million for

prohibited AI practices

-  Up to 3% of global annual turnover or EUR 15 million for

most other violations, including non-compliance with highrisk system requirements

-  Up to 1% of global annual turnover or EUR 7.5 million for

supplying incorrect, incomplete or misleading information

-  Any individual can make non-compliance complaints


**1. Help drive AI risk mitigation**
Engage early to understand the process, data governance,
privacy and other risks associated with AI while safely taking
full advantage of its value and opportunities for your business.
People from diverse backgrounds and roles should all have
a seat at the table when creating an AI strategy that’s
transparent and responsible.

**2. Preparation is key to setting your organization up for success**
We expect that similar legislation will be adopted by many
regulators globally. Here’s some action you should start
thinking about taking now to stay ahead:

– Determine your role(s) in AI value chain as defined

by the EU AI Act, e.g. deployer, distributor etc.

– Create an overview of the AI already in use in your

company, regardless of whether your organization
or a third-party has created it.

– Perform a risk assessment to categorise the AI based

on the categories in the EU AI Act and other risks
relevant to your organization.

– Introduce end-to-end, cross-company AI governance

frameworks to help meet your obligations for existing
and future AI.

– Verify that your suppliers are meeting their obligations,

which may be a requirement for your organization
to meet its obligations.

– Consider extending existing compliance programs

to incorporate new regulatory requirements.

– Establish clear responsibilities and organizational

structures for AI governance and AI ethics,
like an AI Ethics Board.


-----

## pp

IBM can help you prepare for the changing regulatory landscape
by weaving a responsible, transparent AI strategy into your
business and helping you enact that strategy on a day-to-day
basis.

[Request a strategy session](https://www.ibm.com/consulting/ai-governance)

[Explore the AI governance toolkit](https://www.ibm.com/products/watsonx-governance)


© Copyright IBM Corporation 2024. IBM, the IBM logo, IBM Consulting,


**focus on building trust**
An AI strategy that adheres to strong, responsible practices will
ultimately protect people while also promoting innovation in this
exciting space. What should an AI strategy look like?

– Customize approach: Establish different rules for different

risks. Remember, it’s not a one-size-fits-all approach.

– Be transparent: AI systems that directly interact with people

should clearly inform users when they are interacting with
artificial intelligence.

– Be explainable: Be prepared to demonstrate how data is

obtained and how AI is trained.

– Auomate workflows: Doing this whereever possible helps

ensure compliance is not taking more time from development
and other business teams than needed.

**4. Look for AI governance tools that align with your AI strategy**
Ultimately, a 3-pronged approach to AI governance that
includes compliance, risk management and lifecycle
governance can help you to achieve compliance with
the changing regulatory landscape.

[IBM® watsonx.governance™ is an AI governance toolkit](https://www.ibm.com/products/watsonx-governance)
built to address all 3 of these pillars.

Direct, manage and monitor your organization’s AI activities
whether they are created or deployed on an IBM platform,
third-party platform or even embedded in business applications.


-----

