CIOs Lag on EU AI Act Readiness; Training Shows Progress

New findings from the diginomica network reveal that many CIOs and enterprise leaders are falling behind in their preparations for the upcoming EU AI Act Source. While full enforcement is set for December 2027, interim deadlines, such as the August 2026 watermarking requirement, demand immediate attention.

Slow Progress on Key AI Act Requirements

The research highlights concerning statistics regarding compliance. Only 35% of digital leaders surveyed have begun tracking, labeling, or watermarking AI-generated content (images, audio, video, text). Merely 3% have completed this task, with 28% still in progress Source.

Furthermore, a significant portion of businesses has not assessed the Act's applicability. Only a quarter of respondents believe they are in scope of the Act, and 36% had not conducted an assessment to determine if it applies to their operations. This survey was conducted when enforcement was still anticipated to begin in August 2026, underlining the urgency for businesses with European interests Source.

AI Literacy Training Shows Greater Engagement

Despite the slow uptake in other areas, the human-centric aspect of the EU AI Act, specifically requiring regular AI training for staff to enhance AI literacy, is seeing better engagement. Approximately 38% of the community are partially compliant, 34% are actively working towards compliance, and 13% report full compliance in this area Source.

Mark Chillingworth notes that while compliance can be viewed as a hindrance, the Act's focus on people and understanding the implications of technology is a positive development. This emphasis on AI literacy is considered overdue and crucial for navigating technological advancements effectively Source.

Broader Regulatory Landscape and Innovation Considerations

The EU AI Act is just one piece of a growing global regulatory puzzle. Enterprises will encounter regulations from various sources, including US states, federal initiatives, and emerging frameworks in Latin America. The challenge extends beyond compliance to fundamental questions about technical dependence on external AI providers and hyperscalers, as well as the need for sovereign data centers and localized models to foster genuine innovation Source. Getting these strategic decisions right will be essential for guiding innovation rather than merely enforcing a status quo.