Many Indian software companies sell to European clients: SaaS products, custom builds, chatbots, analytics and outsourced development. If any of that includes AI, the EU Artificial Intelligence Act (Regulation (EU) 2024/1689) is now part of your client's procurement checklist, and in some cases your own legal obligation. The deadlines also moved in 2026, so older summaries are out of date.
This EU AI Act summary is written for Indian founders and delivery heads. It covers who the Act reaches outside Europe, the risk tiers, provider and deployer duties, the current dates and a practical action list. It is not legal advice; for a specific product, confirm your classification with EU counsel.
Does the EU AI Act apply to an Indian company?
Yes, in several situations. Article 2, as published on the AI Act Explorer, covers:
- Providers anywhere in the world that place AI systems or general-purpose AI models on the EU market or put them into service there.
- Providers and deployers outside the EU where the output produced by the AI system is used in the EU.
- Importers, distributors and authorised representatives of non-EU providers.
So an Indian company selling an AI-powered SaaS tool to German customers is likely a provider. An Indian BPO using an AI screening tool whose results are used for an EU client may be caught through the output test. A pure development contractor building to the client's specification may or may not be the provider, depending on whose name the system goes to market under. Clarify this in the contract, not after an audit.
The four risk tiers
The European Commission's AI Act overview describes four levels:
| Tier | Examples given by the Commission | What it means |
|---|---|---|
| Unacceptable risk | Social scoring, untargeted scraping of facial images, emotion recognition in workplaces and schools, AI generating non-consensual sexual content or child sexual abuse material | Banned |
| High risk | Safety components in critical infrastructure, exam scoring, CV sorting in recruitment, credit scoring, AI preparing court rulings | Allowed with strict obligations: risk management, data quality, logging, documentation, human oversight, robustness and accuracy |
| Transparency risk | Chatbots, generative AI content, deepfakes | Users must be told they are dealing with AI; AI-generated content must be identifiable; deepfakes labelled |
| Minimal risk | Spam filters, AI in video games | No specific rules |
Separate rules apply to providers of general-purpose AI models, which matters mostly to companies training large foundation models rather than building on top of them.
Where Indian exporters usually land
- HR tech: resume screening, candidate ranking and employee performance tools are typical high-risk cases.
- Fintech and insurance: creditworthiness scoring for individuals is high risk.
- Edtech: AI that scores exams or decides admissions is high risk.
- Customer service chatbots, content and marketing tools: usually transparency obligations.
- Internal analytics, forecasting and spam filtering: often minimal risk.
Provider vs deployer: who does what?
For high-risk systems, the Act splits duties between the provider, who develops the system and places it on the market under its name, and the deployer, who uses it professionally.
Provider duties (Article 16)
Under Article 16, providers of high-risk systems must, among other things, meet the technical requirements, run a quality management system, keep technical documentation and logs, complete a conformity assessment, issue an EU declaration of conformity, affix the CE marking, register the system where required, take corrective action when problems appear, and demonstrate conformity to authorities on request.
Non-EU providers of high-risk systems must also appoint an authorised representative established in the EU by written mandate before placing the system on the EU market, under Article 22. The representative keeps the declaration and technical documentation available for 10 years and deals with authorities.
Deployer duties (Article 26)
Under Article 26, deployers must use the system according to its instructions, assign competent human oversight, ensure input data they control is relevant and representative, monitor the system and report risks and serious incidents, keep automatically generated logs for at least six months, inform workers before workplace use, and inform people when an Annex III system makes or assists decisions about them.
Your European client is usually the deployer, which is why they will ask you, the provider, for the documentation and instructions they need to meet their own duties.
EU AI Act dates after the 2026 Digital Omnibus
The Act entered into force on 1 August 2024. In 2026 the EU adopted the Digital Omnibus on AI, Regulation (EU) 2026/1744, which entered into force on 27 July 2026 and pushed back the high-risk dates, as recorded by the AI Act Explorer and the Commission's own overview. The current timeline:
| Date | What applies |
|---|---|
| 2 February 2025 | Original prohibited practices and AI literacy obligations |
| 2 August 2025 | Governance rules and obligations for general-purpose AI models |
| 2 August 2026 | Act generally applicable; transparency rules such as chatbot disclosure apply |
| 2 December 2026 | New ban on AI generating non-consensual sexual content or CSAM; end of grace period for content-marking duties for generative systems already on the market |
| 2 December 2027 | High-risk rules for Annex III use cases such as employment, education, biometrics and credit |
| 2 August 2028 | High-risk rules for AI embedded in regulated products under Annex I, such as machinery, lifts and toys |
The omnibus also softened the AI literacy duty, so providers and deployers must now take measures to support AI literacy among staff rather than ensure a sufficient level of it, and extended some simplified requirements to small mid-cap companies. Check the consolidated text before relying on any single date for your product.
What are the penalties?
Under Article 99, maximum fines are up to €35 million or 7% of worldwide annual turnover for prohibited practices; up to €15 million or 3% for breaches of other operator obligations, including transparency; and up to €7.5 million or 1% for supplying incorrect or misleading information to authorities. For each tier the higher figure applies, except that for SMEs and start-ups the lower figure is the cap.
Action list for Indian software exporters
- Inventory your AI. List every product and service with an AI component sold into or used in the EU, including third-party models you embed.
- Classify each one against the tiers above, and record your reasoning.
- Decide your role per contract: provider, deployer, distributor or contractor. Put it in writing with the client.
- Fix transparency now. Chatbot disclosures and AI-content marking apply already. These are usually quick wins.
- For high-risk products, start documentation early. Risk management, data governance, logging, human oversight design and a quality management system take months, not weeks. December 2027 is closer than it looks.
- Plan the EU authorised representative if you will be a non-EU provider of a high-risk system.
- Prepare for client questionnaires. Expect European buyers to ask for your AI Act position, model documentation and incident process during procurement.
- Train your team. AI literacy remains an obligation, and it is good practice anyway.
How Cehpoint can help
Our EU AI Act compliance and AI governance service covers AI inventory, risk-tier classification, gap analysis and the technical documentation European clients and auditors expect. Where security evidence is also required, our ISO 27001 and SOC 2 readiness work fits alongside, and our AI red teaming service provides the robustness testing evidence high-risk systems need.