EU AI Act: How can Businesses Ensure Compliance?

The newly passed EU AI Act has attracted considerable interest from business leaders and policymakers, according to Boston Consulting Group (BCG).

With the new regulations expected to be enforced from May 2024, businesses now have time to consider how to best comply with and enforce the new rules. During an era of digital transformation, having rules in place concerning AI will be beneficial to avoid malicious use, whilst also informing businesses on the most appropriate use cases.

AI regulations are vital for businesses, as it allows them to be aware of the potential risks of algorithmic bias, data privacy concerns and cyber disruption like a breach.

BCG recently released insights in collaboration with the Responsible AI Institute to advise business leaders how to best implement AI standards across their organisations. AI Magazine considers these steps alongside the global business landscape.

“Set objectives for using AI standards”

AI standards can be used to strengthen trust in AI, as well as wider acceptance within various key industries and wider society. Standards for AI are beneficial for the quality, security, traceability, transparency and reliability of AI applications.

Business leaders can help shape the AI governance landscape by developing clear and direct AI standards for their workforce. By implementing lessons learned and gaining access to best practices, companies can contribute to standardisation at a national or international scale.

These standards also enable organisations to consider social and environmental factors in AI standards.

“Select AI standards”

Selecting AI standards can offer businesses a competitive edge by promoting responsible AI use, streamlining operations and reducing risks. The benefits afforded by these standards can provide practical guidance to those within an organisation and promote greater innovation.

BCG and the Responsible AI Institute advise that companies would benefit from focusing on well-known AI standards first, such as ISO/IEC 42001 and ISO/IEC 23894. 

“These standards will have more robust supporting communities and their terms and concepts may be incorporated by suppliers, customers, end-users and government organisations. In many cases these are also certifiable and auditable,” the organisations said.

“Identify the standards your organisation uses and pinpoint expertise”

BCG and the Responsible AI Institute highlight these functions can serve as sources of expertise and pilot candidates for introducing new standards. They cite cybersecurity and risk management teams as valuable sources of information on how to best adapt and conform to new AI standards.

Organisations can start this by reviewing all relevant documentation concerning its AI models and consider existing processes. This could lead to internal audits, such as surveys, self-assessments and practical tests, to identify expertise within specific teams to improve standards and AI methodologies. 

“Conduct a gap analysis to identify next steps”

Conducting analysis will allow businesses to prioritise gaps within their technology based on your organisation’s objectives.

“Map the selected AI standards to your organisation’s existing AI strategy, governance, and processes and identify gaps,” both organisations suggest. “Determine importance and estimated cost for each gap.”

“Engage in standards processes”

Being able to contribute to implementing and maintaining AI standards will inevitably put a business ahead, as they are committing to shaping the AI governance landscape. Businesses can get involved via national committees or with standards organisations directly, in order to remain at the forefront of emerging best practices when it comes to new technologies.

AI is here to stay and so organisations will benefit by embracing it safely.

************

Make sure you check out the latest edition of AI Magazine and also sign up to our global conference series - Tech & AI LIVE 2024

************

AI Magazine is a BizClik brand