Advertisement

Explainer: What the EU AI Act’s transparency rules mean for Kenya’s tech firms

Explainer: What the EU AI Act’s transparency rules mean for Kenya’s tech firms
Safaricom Headquarters in Nairobi.PHOTO/@SafaricomPLC/X

Kenyan technology companies, digital agencies, fintech firms and media organisations serving customers in the European Union (EU) face new compliance obligations from August 2, 2026, under the EU AI Act.

Businesses that fail to meet the law’s AI transparency requirements could face fines of about Ksh2.2 billion (€15 million) or 3 per cent of their global annual turnover, whichever is higher.

The new rules, introduced under Article 50 of the EU AI Act, are designed to increase transparency around artificial intelligence, curb the misuse of deepfakes and help consumers identify AI-generated content. For Kenyan firms with European clients or users, understanding the law is becoming a business necessity rather than a regulatory formality.

What is Article 50 of the EU AI Act?

Article 50 sets out transparency obligations for AI-generated content. Its core requirement is that AI-generated material should be identifiable, even when it closely resembles content created by humans.

President William Ruto and European Union President Ursula von der Leyen.PHOTO/@WilliamsRuto/X

Companies that develop generative AI models must ensure the content their systems produce, including images, videos, audio recordings and synthetic voices, contains machine-readable markers or other technical measures that enable detection by digital systems.

These identifiers may not be visible to users but allow platforms and regulators to recognise AI-generated content.

The objective is to improve trust in AI while making it harder for manipulated or synthetic media to be presented as authentic.

Who must comply?

The law distinguishes between developers of AI models and businesses that deploy AI tools.

European Parliament during a past session.PHOTO/@Europarl_EN/X

The most demanding technical obligations, including watermarking and AI detection measures, fall on companies developing generative AI systems. Businesses using AI tools such as ChatGPT, Midjourney or similar platforms generally are not expected to build these technical systems themselves.

However, deployers still have important responsibilities. Businesses, marketers, media organisations and other users must clearly disclose when publishing AI-generated or AI-manipulated content on matters of public interest. Deepfakes must be labelled accordingly, while customer service chatbots must inform users that they are interacting with artificial intelligence unless that fact is already obvious.

Effects on Kenyan businesses

The financial penalties are significant. Companies that breach the transparency rules can be fined up to €15 million or 3 per cent of their worldwide annual turnover, whichever is higher. For large multinational companies, turnover-based penalties could substantially exceed the fixed fine.

While the rules may present a compliance challenge for smaller AI startups and independent content creators, the greatest technical burden remains with AI developers rather than ordinary business users.

Inside office branded with Safaricom colours and logos. PHOTO/https://www.facebook.com/SafaricomPLC
Inside office branded with Safaricom colours and logos. PHOTO/https://www.facebook.com/SafaricomPLC

The rules apply to companies operating in or serving the EU market, regardless of where they are based. That means Kenyan marketing agencies producing AI-assisted campaigns for European clients, fintech platforms with EU users, software companies, outsourcing firms and media organisations syndicating AI-assisted content to European partners may all be affected.

Businesses should identify where AI is used across their operations, establish clear disclosure practices and review contracts with technology providers to ensure compliance. Organisations developing their own AI models should also assess whether their systems meet the Act’s technical transparency requirements.

There may be additional time for some companies. A provisional agreement reached by EU lawmakers in May 2026 under the Digital Omnibus package would allow providers already operating in the EU market before August 2026 until December 2026 to fully comply, easing the transition for existing businesses.

For Kenyan companies with European customers, early preparation will be critical. As the EU AI Act begins reshaping global standards for AI governance, firms that strengthen transparency and compliance now will be better positioned to maintain access to one of the world’s largest and most tightly regulated digital markets.

Author

For these and more credible stories, join our revamped Telegram and WhatsApp channels.
Advertisement