If the company you work for or your organisation uses Artificial Intelligence-based tools every day, from Generative AI software for marketing and reporting to customer service chatbots, copilot systems built into company CRMs or recruiting tools, there is a piece of legislation every business should know about and a date it should mark in red on its calendar: 2 August 2026.
You’ve probably noticed how, more and more often in radio adverts or traditional media, announcements specify with dedicated disclaimers that “the voice in this advert was generated using Artificial Intelligence”. This isn’t a communication quirk or a marketing trend, but the direct anticipation of what European legislation already requires and will extend across the board: radical transparency about algorithm-generated content.
This is, in fact, when the operational obligations on transparency, traceability and labelling laid down by Article 50 of Regulation (EU) 2024/1689 (AI Act) become fully applicable.
Established to ensure transparent, safe and human-centric technological innovation, the European Regulation sets clear boundaries for anyone who develops or uses Artificial Intelligence algorithms. Being caught unprepared on 2 August 2026 exposes a business to financial penalties, as well as real risks of its technologies being blocked and serious damage to its corporate reputation.
Adopting robust AI governance early is not just about avoiding penalties: it protects your processes, enhances your brand and turns compliance into a clear competitive advantage in the market.
AI Act: the key milestones
The AI Act phases in its requirements gradually, giving organisations time to adapt step by step:
- 1 August 2024: Regulation (EU) 2024/1689 formally enters into force.
- 2 February 2025: The bans on unacceptable-risk systems (e.g. social scoring, subliminal behavioural manipulation) take full effect, along with the AI literacy obligations (training staff across the organisation).
- 2 August 2025: Full application of the rules on general-purpose AI models (General Purpose AI – GPAI).
- 2 August 2026: Article 50 of the AI Act comes into application, covering direct obligations on transparency, content traceability and digital labelling for Providers and Deployers. From this moment the penalty regime officially kicks in, with inspections and administrative fines enforceable by the Supervisory Authorities.
- Looking ahead to 2027–2028: Full compliance required for systems classified as high-risk (e.g. biometrics, human resources management, credit) and for embedded products subject to sector-specific legislation (Annexes I and III).
What the AI Act Is and What the European Legislation Requires
The AI Act (EU Regulation 2024/1689) is the world’s first comprehensive, binding legal framework created to regulate the development, placing on the market and use of Artificial Intelligence within the European Union.
With this law, Europe aims to strike the right balance between accelerating technological innovation and protecting fundamental rights, as well as citizens’ safety and privacy. It is not simply a set of ethical recommendations, but a directly applicable EU regulation that imposes clear rules, traceable responsibilities and rigorous compliance standards on all private and public organisations.
1. Classifying AI systems into 4 risk levels
The Regulation does not apply the same rules to every technology; it scales them according to the potential harm to people and society:
- Unacceptable risk (prohibited practices): Systems posing a clear threat, which are banned outright (e.g. social scoring systems, subliminal behavioural manipulation, biometric categorisation based on sensitive data or emotion recognition in the workplace).
- High Risk: Software used in critical contexts (e.g. recruitment and personnel management, credit scoring, critical infrastructure, healthcare, access to public services). These systems require a complex conformity procedure: fundamental rights impact assessment, risk management, training data quality, traceability and ongoing human oversight.
- Specific Transparency Risk (Medium Risk): Artificial Intelligence systems that people interact with directly or that generate synthetic content (e.g. chatbots, business copilots, Generative AI, algorithms for producing deepfake content). For these systems, the strict transparency, disclosure and labelling obligations set out in Article 50 apply.
- Minimal or no risk: Most traditional AI applications (e.g. spam filters, simple recommendation algorithms, video games), for which only voluntary codes of conduct are recommended.
2. Who does it apply to? Providers and deployers
The AI Act precisely defines the roles along the technology’s value chain, assigning targeted responsibilities:
- Providers: Anyone who develops, customises or has developed an AI system (or a general-purpose AI model – GPAI) in order to place it on the market or put it into service under their own name or trademark.
- Deployers (Users): Any natural or legal person, public authority or company that uses an AI system under its own authority in the course of a professional activity. (Note: your company is a Deployer if it uses third-party AI software for marketing, HR, sales or business process optimisation).
3. The cross-cutting “AI Literacy” obligation (Article 4)
Often underestimated but already in force, Article 4 of the AI Act requires all businesses that use AI systems to ensure an adequate level of Artificial Intelligence literacy (AI Literacy) among their staff. This means employees and collaborators must be properly trained and educated on the correct functioning, limitations, risks and ethical, safe use of the algorithmic technologies adopted in the company.
From the AI Act to AI Governance: how companies’ approach to using Artificial Intelligence is changing
The AI Act is not simply a set of regulatory constraints: it anticipates and accelerates what has already become a strategic priority for the most forward-thinking companies, namely AI Governance.
While many organisations see Artificial Intelligence solely as a practical tool for speeding up processes, market leaders have grasped a fundamental rule: AI cannot be integrated into a company in an unregulated or uncontrolled way.
What is AI Governance, in simple terms?
In practical terms, AI Governance is the set of rules, internal processes, roles, responsibilities and technological tools that a company defines to direct, monitor and control how Artificial Intelligence is developed, purchased or used by employees and IT systems.
Doing AI governance means answering key questions such as:
- Which AI software are we buying, and who authorises it?
- Are our employees entering confidential company data or trade secrets into ChatGPT or other generative tools?
- Are the outputs generated by algorithms (e.g. contracts, marketing campaigns, HR selections) checked by a person before they take effect?
- Do the AI software suppliers we use comply with the security and transparency standards set by the European Union?
What are the risks of NOT applying AI governance in your company?
Operating without an AI governance model means driving with your headlights off through terrain full of legal, financial and reputational pitfalls:
- 1. Severe financial penalties: Failure to comply with the transparency, labelling or risk management obligations under the AI Act and the GDPR exposes a business to fines of up to €15 million or up to 3% of total worldwide annual turnover.
- 2. Data and know-how loss (data leakage): Without clear rules, employees tend to use free online tools on their own initiative (known as shadow AI). Entering financial data, accounts, code or commercial strategies into external software without proper contractual safeguards means handing them over to third-party training algorithms.
- 3. Reputational damage and legal liability: Publishing incorrect information (AI hallucinations), unauthorised images or algorithm-generated text without genuine human review (Human-in-the-Loop) can lead to legal disputes and destroy the brand’s credibility in an instant.
- 4. Operational shutdown and wasted resources: Investing significant budgets in software or integrations that do not comply with European legislation carries the real risk of an emergency suspension by the Supervisory Authorities, forcing the company to halt its activities or redesign everything at very high cost.
On the contrary, managing AI with method and awareness isn’t just about avoiding fines: it protects your information assets, reassures clients and partners and lets you gain a real competitive advantage, using technology safely and effectively ahead of the competition.
How can we support you with corporate AI Governance?
Complying with the AI Act and managing Artificial Intelligence strategically should not be seen as a bureaucratic brake, but as an opportunity to reorganise and safeguard your company’s processes.
Studio Salatino supports business owners and managers in defining tailored practices, methodologies and policies for corporate AI governance.
What we do:
- AI Audit & Mapping: We analyse the software and algorithmic tools already in use to identify any legal, contractual or security issues at an early stage.
- Defining Policies and Processes: We set up clear, simple and workable internal rules for the correct use of AI by employees and collaborators.
- Supplier Compliance Checks: We guide you in reviewing third-party software, making sure it meets the transparency and marking requirements set by the legislation.
- Training and AI Literacy: We design AI literacy programmes for staff, in line with the obligations set out in Article 4 of the Regulation.
With our multidisciplinary support, you can mitigate risks, eliminate the danger of fines and make the most of Artificial Intelligence’s full potential to grow your business with complete peace of mind.
Don’t wait for the 2 August 2026 deadline to make your organisation secure.
Contact us for information about our AI governance services
Please note that the content published is for general information purposes only and does not in any way constitute legal advice or professional consultancy replacing the opinion of a qualified lawyer. Legislation on Artificial Intelligence is evolving and requires an in-depth, case-by-case analysis; we therefore recommend always consulting the full official text of EU Regulation 2024/1689 (AI Act) and the relevant regulatory documentation to check the details of the law. This article was written with the help of Artificial Intelligence tools and was fully checked, verified and editorially reviewed by a human professional.
