EN

Software suppliers · Art. 25 · deployer duties

Vendor AI: the AI you did not buy as AI, but may still be responsible for

The most sensitive regulatory form of shadow AI often sits in software that has been in use for years and has received new AI components through regular updates.

What vendor AI is

When financial institutions think about artificial intelligence, they usually think of conscious decisions: buying a new AI tool, starting a pilot project, approving a budget. Reality is often different. AI may already be part of the existing software landscape.

This is vendor AI: AI functionality in systems from external providers. For banks and insurers it is critical because deployer duties may arise even if the institution did not develop the model itself.

Typical vendor AI systems

AreaTypical providersEU AI Act relevance
Core banking / lendingTemenos, Finastra, MambuAI support in credit decision processes may touch Annex III No. 5(b).
Credit scoringFICO, Moody's Analytics, SASCreditworthiness assessment of natural persons is a direct high-risk area.
AML / fraud detectionNICE Actimize, Oracle FCCMFraud detection is expressly excluded from Annex III No. 5(b); other functions may still trigger data-protection, DORA or governance duties.
Life/health underwritingGuidewire, Sapiens, Duck CreekRisk assessment and pricing in life/health insurance can be high-risk AI.
HR recruitingPersonio, SAP SuccessFactors, WorkdayRecruiting, matching, performance evaluation and monitoring may trigger high-risk AI and labour-law duties.
Chatbot / customer serviceGenesys, NICE CXone, Leena AITransparency duties under Art. 50 and data-protection issues in customer dialogues.

The role shift under Art. 25

Art. 25 EU AI Act contains a practically important rule: an organisation that operates a purchased AI system under its own name, substantially modifies it, or uses it for a different purpose can move into provider obligations. A deployer role then becomes a significantly more demanding compliance situation.

  • Branding: operation of a vendor system under the institution's own name or brand.
  • Substantial modification: adaptation affecting purpose, performance or risk classification.
  • Change of purpose: use for a purpose different from that intended by the provider.

The assessment depends on the individual case. A wrong assumption that the institution is only a deployer can create significant gaps in documentation, conformity and supplier control.

CSRD / ESRS: vendor AI as part of the value chain

Vendor AI can also be relevant for CSRD / ESRS. CSRD requires reporting on own operations and, where applicable, value chains, products, services and business relationships. If purchased AI systems cause discrimination risks or human-rights impacts, procurement, supplier management and governance need to be prepared.

What a systematic vendor AI analysis must cover

The vendor list must not only be searched for product names. Release notes, contractual basis, actual use, business processes, role allocation and risk classification are decisive. Procurement, IT, compliance and business units must work together.

Identify vendor AI systematically

We support the systematic identification and classification of vendor AI in your software landscape.

Contact us
Notice: This page is for general information only and does not constitute legal advice. Role allocation under Art. 25 EU AI Act must be legally assessed case by case.