An analysis of the EU AI Act and the May 2026 political agreement on the Digital Omnibus on AI

On 7 May 2026, the European Council and the European Parliament reached provisional political agreement on the Digital Omnibus on AI, revising key compliance timelines under the EU AI Act. The Act’s obligations for the e-mobility sector, however, still apply. The agreed package delays certain high-risk AI requirements affecting charging infrastructure and fleet management. It also revises the deadline for AI content labelling. Core provisions, however, are already in force, and the sector transformation is well underway. This article sets out what e-mobility legal and compliance teams need to know: which timelines may shift, which obligations apply today, and why continued investment in AI governance remains both a legal and a competitive priority.

Agreed delays: what the AI Omnibus actually says

On 19 November 2025, the European Commission published its Digital Package, including the Digital Omnibus on AI Regulation (the “AI Omnibus Proposal”). Until that announcement, 2 August 2026 stood as the enforcement milestone for high-risk AI obligations. On 7 May 2026, the Council and the Parliament reached provisional political agreement on the Digital Omnibus on AI, confirming a revised set of deadlines. The agreed package introduces conditional delays for high-risk obligations and a revised deadline for AI content labelling, but it does not remove the underlying obligations. Formal adoption and publication in the Official Journal are still required before the new timelines become binding law.

Under the agreed package, full application of high-risk AI obligations would be deferred until the Commission confirms that adequate compliance support infrastructure — harmonised standards, common specifications, and official guidelines — is available. Once that confirmation is issued, transition periods would apply as follows:

For AI content labelling under Article 50, the agreed package moves the deadline from 2 August 2026 to 2 December 2026, and reduces the granted grace period for provider to a three-month period. All other generative AI obligations remain unchanged. Providers and deployers of generative AI should plan to implement labelling for AI-generated content in customer-facing outputs by 2 December 2026.

AI systems in EV charging networks, smart grid management, fleet optimisation, driver monitoring, and AI-powered customer service mostly fall within the scope of Annex I or Annex III. The proposed delays extend deadlines, but they do not change which systems fall in scope. CPOs, MSPs, OEMs and fleet operators must begin classification now.

Why governance cannot wait: three reasons

1. The political agreement is not yet binding law

Although the Council and Parliament reached provisional political agreement on 7 May, the Digital Omnibus on AI still requires formal adoption and publication in the Official Journal before the revised timelines become binding law. Until formal adoption, the original AI Act timelines remain in force. Compliance teams can use the agreement as a planning baseline from today, while continuing formal preparation against the AI Act’s current timelines.

2. Core obligations are already binding

The EU AI Act entered into force on 1 August 2024. The following obligations apply today, regardless of any Omnibus outcome:

• Prohibited AI practices, including social scoring, manipulative AI, real-time biometric surveillance in public spaces, and, following the May 2026 agreement, the generation of non-consensual sexual imagery and child sexual abuse material.
• Transparency obligations for AI systems interacting with natural persons.
• Full compliance framework for general-purpose AI (GPAI) models, including systemic risk obligations for frontier models.

This is current law, not future requirements. Any e-mobility company deploying AI in customer-facing applications, chatbots, or content generation must already comply with.

3. Governance is a market expectation, not just a legal one

Even where the Omnibus Proposal extends technical deadlines, it does not reduce the underlying obligations. Accountability, transparency, human oversight, and valid documentation are now market expectations — held by customers, public procurement authorities, investors, and business partners alike. Companies that treat this period as a pause rather than a preparation phase will reach enforcement deadlines without the foundations in place. Acting now is more efficient than catching up later.

How the EU AI Act reshapes e-mobility: key compliance domains

Smart charging and grid integration

AI systems optimising energy distribution, load balancing, and demand prediction are likely classified as high-risk under Annex I when embedded in regulated products or critical infrastructure. Obligations include risk management systems throughout the AI lifecycle, representative and bias-free training datasets, detailed technical documentation, human oversight mechanisms enabling operator intervention, and cybersecurity with automatic event logging. Compliance will extend time-to-market, but will also become a differentiator in public tenders and enterprise contracts.

Autonomous and connected vehicle systems

AI in autonomous vehicles, ADAS, and V2X platforms fall within the high-risk framework under the Motor Vehicle Regulation. Mandatory requirements include safety validation under diverse conditions encompassing adversarial edge cases, transparency obligations on AI versus human control, continuous post-market performance monitoring, and incident reporting to competent authorities. OEMs and suppliers will need to embed AI governance into their development processes from the design phase onwards.

Fleet management and employment AI

AI affecting driver monitoring, route optimisation with employment implications, and workforce allocation is classified as high-risk under Annex III. Obligations include detecting and preventing discriminatory outcomes, informing workers about AI systems affecting their employment, ensuring meaningful human review rights for automated decisions, and aligning AI Act compliance with GDPR obligations on automated decision-making. Fleet operators will need to invest in explainable AI and human-in-the-loop mechanisms, working closely with HR, legal, and labour representatives.

Customer-facing and generative AI

Chatbots, recommendation engines, and generative AI tools are subject to transparency obligations that are already in force: users must be informed they are interacting with an AI system, and AI-generated content must be labelled as such. Following the May 2026 political agreement, the Article 50 deadline for AI content labelling has been moved from 2 August 2026 to 2 December 2026. Companies deploying these tools will need to implement disclosure mechanisms that naturally fit into the user’s experience. Beyond compliance, transparency also strengthens customer trust.

Supply chain and procurement

The Act assigns obligations across the AI value chain. E-mobility companies integrating third-party AI components are responsible for their suppliers’ compliance. This requires structured vendor due diligence, contractual provisions mandating compliance and documentation handover, a clear allocation of provider versus deployer responsibilities, and standardised audit protocols. Legal and procurement teams will need to develop AI-specific supplier assessment frameworks before deadlines approach.

Priority actions for 2026: building governance that lasts

Map your AI use cases. Cover charging hardware, fleet platforms, predictive maintenance tools, and customer systems. Build a centralised inventory with risk classification under Annex I and Annex III.

Assess high-risk status. Pay particular attention to systems affecting critical infrastructure, employment decisions, and vehicle or charging equipment safety.

Set up internal governance. Define oversight roles, AI development and deployment policies, transparency, data quality controls, and incident response protocols.

Strengthen supplier oversight. Develop AI-specific due diligence questionnaires and audit mechanisms. Ensure contracts require compliance with documentation and cooperation with regulatory obligations.

Companies that build strong AI governance now will be best placed to meet regulatory developments, satisfy stakeholder expectations, and compete in a market where AI accountability is no longer optional.

Ilenia Lombardo
Head of Legal & DPO
Last Mile Solutions