Skip to Content
Withdraw from contract

Curriculum & Assessment Programme

KI-Pass – six-month AI skills programme with specialist practice

Move from responsible AI use to evidence-based application at work. The KI-Pass combines a shared foundation, a chosen specialist track, supported practical experience and a competence-based final assessment. The outcome records what you can demonstrate, rather than attendance alone.

ELA-PRG-KIPRevision 1.0 · 31 October 2026
KI-Pass – six-month AI skills programme with specialist practice

Audience and entry requirements

Who is this programme for?

Working professionals, career changers and teams who want to use AI practically or support its responsible deployment. The specialist track is chosen according to existing knowledge, the intended role and a suitable practical assignment.

What are the entry requirements?

A structured entry discussion reviews digital foundations, learning goals, available study time and prerequisites for the chosen track. Teaching and assessment are in German or English as agreed; sufficient command of the selected language is needed for assignments and the professional discussion. No particular university degree is required. The development track requires additional programming and data-handling knowledge, which may be demonstrated through an agreed preparatory plan. The practical route must receive technical approval before the specialist practice phase begins.

Scope and learning format

Standard model: approximately six months or 26 weeks. It comprises five weeks of shared foundations, three weeks of specialist preparation, 16 weeks of supported specialist practice and a two-week completion phase. The calendar, teaching units, practice hours, supervision, absence arrangements and assessment dates are agreed before booking. The 26 weeks are a newly designed programme model, not a statement about the scope of the separately offered foundation package.

5 weeks

1. Shared Foundations

A shared vocabulary, practical AI use and the agreed introduction to AI management.

3 weeks

2. Specialist Preparation

Specialist learning and approval of a realistic practice and learning plan.

16 weeks

3. Specialist Practice

Sustained application with technical supervision and documented milestones.

2 weeks

4. Completion

Portfolio, reflection and competence-based final assessment.

Which competencies will you demonstrate?

  • Define a professional AI use case with an objective, boundaries and testable success criteria.
  • Use AI tools in a controlled way and assess outputs against traceable standards.
  • Document data provenance, sources, uncertainty and approvals in personal work.
  • Plan, carry out and evidence a bounded assignment in the chosen specialist track.
  • Recognise risks, unresolved questions and limitations and refer them to responsible roles.
  • Explain personal work independently and derive technically defensible improvements.

The curriculum

1. AI Foundations & Compliance

Capabilities and limits of current AI systems, suitable use cases, human responsibility, data and source evaluation, and introductory issues in privacy, information security and organisational approval.

Scope / allocation

Within the five-week shared foundation

Evidence and output

A reasoned use-case choice with documented boundaries and unresolved approval questions.

2. Operational Prompt Engineering

Define tasks precisely, select context and data, specify output formats, test variants, verify results and develop reusable work instructions.

Scope / allocation

Within the five-week shared foundation

Evidence and output

A prompt and test record with a traceable comparison of variants and analysis of errors.

3. Multimodal AI & Data Analysis

Text, image and document processing and structured data work using approved material; interpretation, traceability, quality criteria and limits of automated analysis.

Scope / allocation

Within the five-week shared foundation

Evidence and output

A traceable analysis of a dataset or document collection with checked claims and sources.

4. AI Management with ISO/IEC 42001

Orientation to the PECB ISO/IEC 42001 Foundation and Lead Implementer learning routes; requirements, roles, risks, documented processes and improvement of an AI management system. External PECB examination preparation follows the specifically agreed scope.

Scope / allocation

Learning units integrated into the foundation phase under the agreed schedule

Evidence and output

Case-based allocation of responsibilities and management-system tasks; external PECB examinations remain separate.

5. Specialist Track and Practical Assignment

Specialist preparation, baseline review, concrete learning goals, practical-project boundaries, acceptance criteria and agreement with the technical supervisor.

Scope / allocation

Three weeks of specialist preparation in the standard model

Evidence and output

An approved practical learning plan with three concrete project outputs and four milestones.

6. Supported Specialist Practice

Independent application in a workplace or technically approved practice environment; documented work, regular feedback, interim reviews and reflective treatment of errors and changes.

Scope / allocation

16 weeks in the standard model; binding practice hours specified in the individual plan

Evidence and output

Weekly journal, four milestone reviews and the three track-specific project outputs.

7. Completion and Evidence of Competence

Prepare practical results, reflect on limitations and alternatives, complete a new practical task and defend decisions in a professional discussion.

Scope / allocation

Two-week completion phase in the standard model

Evidence and output

Assessed practice portfolio, written knowledge component, practical task and presentation with discussion.

Choose your AI Pass specialisation

AI Application Development

For aspiring AI application developers and professionals who want to build credible AI prototypes and integrations. Basic programming and data-handling knowledge is checked before this track begins; a preparatory learning plan can be agreed.

  • Translate a use case into testable technical requirements and a justified AI architecture.
  • Integrate data and interfaces into a controlled AI prototype.
  • Build a test set and compare model or prompt variants using traceable criteria.
  • Recognise errors, security boundaries and uncertainty and address them with suitable technical or organisational measures.
  • Document the development state reproducibly and give a reasoned recommendation on further use.

The curriculum

  • Requirements and architecture: problem, users, data flows, interfaces, model selection, approval boundaries and technical acceptance criteria.
  • Integrating AI functionality: APIs, structured inputs and outputs, retrieval-augmented generation, tool calls and error handling in a controlled prototype.
  • Systematic quality assessment: test datasets, traceable evaluation rules, incorrect answers, outliers, input attacks, costs and latency.
  • Preparing handover: version control, configuration, logging, documentation, human fallback and limits on production approval.

Specialisation project

A reproducibly tested AI assistant for a bounded use case, such as finding and summarising approved knowledge documents. The prototype must show sources, handle uncertainty and respect defined limits. Production deployment is part of the assignment only where separately and expressly authorised.

  • Requirements and architecture package with data flows, interfaces and acceptance criteria.
  • Reproducible prototype with versions, configuration, test cases and recorded results.
  • Handover and operating proposal addressing risks, use limits, monitoring and human fallback.

AI Project Leadership

For project leaders and subject owners who want to lead AI initiatives from initial scoping to an informed decision on piloting and adoption.

  • Define an AI project mandate with testable objectives, boundaries and benefit assumptions.
  • Turn dependencies between data, technology, people and organisation into a workable plan.
  • Manage risks and open decisions transparently and refer them to the responsible roles.
  • Evaluate pilot findings against acceptance criteria agreed in advance.
  • Defend a reasoned recommendation to adopt, improve or stop the initiative.

The curriculum

  • Mandate and benefit hypothesis: stakeholders, problem definition, baselines, achievable outcomes, boundaries and success criteria.
  • Delivery and control: work structure, resources, data dependencies, procurement, budget assumptions and decision points.
  • Piloting and acceptance: testing, user feedback, operational implications, quality criteria and competing objectives.
  • Adoption and handover: change, learning needs, responsibilities, decision papers and follow-up of expected benefits.

Specialisation project

Plan and support a bounded AI pilot, for example an internal knowledge assistant or assisted document processing. Assessment concerns the quality of the project decision, including a defensible recommendation not to proceed.

  • Project mandate with measurable objectives, benefit assumptions, stakeholders and bounded scope.
  • Pilot and control package covering schedule, resource assumptions, risks, decisions and testing.
  • Evidence-based decision paper recommending adoption, further work or closure, with a handover plan.

AI Management & Process Design

For process owners, organisational development specialists and professionals who want to integrate AI into existing workflows and establish traceable operating arrangements.

  • Assess a workflow for suitable AI support and necessary human decisions.
  • Design a future process with clear responsibilities, interfaces and exception handling.
  • Run a controlled workflow prototype or traceable process simulation.
  • Evaluate effects using appropriate process and quality measures without inventing benefits.
  • Write understandable operating and improvement rules for the process.

The curriculum

  • Investigating workflows: current process, handovers, disconnected systems, decision points, data needs and sensible automation boundaries.
  • Designing future processes: allocation of human and AI tasks, approvals, interfaces, responsibilities and exceptions.
  • Tools and implementation: low-code/no-code or rule-based workflow components, controlled tests, documentation and user guidance.
  • Operation and improvement: process measures, feedback, failures, change, training and systematic review of effects.

Specialisation project

Redesign a bounded process using AI support, such as structured preliminary handling of incoming enquiries. The solution must address human approvals, exceptions and a route back to the previous process.

  • Current-state analysis and future process with tasks, data flows, responsibilities and defined approvals.
  • Tested workflow or traceable process simulation with a test record and documented exception cases.
  • Operating and improvement plan with measures, work instructions, learning needs and fallback procedures.

AI Governance & Compliance

For governance, compliance, data protection and management-system professionals who want to document, evaluate and support AI use systematically. The track does not replace legal education or a legally required appointment to a role.

  • Document an AI use case so that purpose, boundaries, responsibilities and data flows can be traced.
  • Assess requirements and risks systematically using identified sources and the actual context of use.
  • Design suitable organisational controls and evidence and assign unresolved technical or legal questions.
  • Critically assess the evidence chain of a use case through a documented review.
  • Explain improvements, approval reservations and remaining uncertainties to management.

The curriculum

  • Inventory and roles: record AI use cases, describe purpose and context, and document responsibilities and decision routes.
  • Risks and requirements: assess sources traceably, consider impacts and data flows, include information security and privacy, and escalate unresolved legal questions appropriately.
  • Controls and evidence: design workable policies, approvals, procurement questions, records, incident routes and monitoring criteria.
  • Management systems and review: understand AI management-system requirements, examine evidence chains, document gaps and prioritise improvement.

Specialisation project

Develop a governance package for a specifically bounded AI use case, such as an internal assistant. It should establish traceable roles, risk treatment, approval criteria and verifiable evidence. A blanket assurance of legal compliance is not a project outcome.

  • Use-case and responsibility file identifying purpose, system boundaries, data flows, ownership and source dates.
  • Reasoned risk and control matrix with open questions, responsible approval roles and required evidence.
  • Governance and review package containing approval arrangements, incident handling, a review checklist and prioritised improvements.

Turning practice into verifiable learning progress

Specialist practice may take place in a suitable organisation, the participant’s own professional setting or a project environment technically approved by the Academy. A workplace placement requires actual work in an organisation with suitable supervision. A simulation or Academy project environment is expressly recorded as such and is not certified as a workplace placement.

A workplace placement is not automatically guaranteed. The location, supervising role and practical arrangements are agreed in writing before practice begins. If no placement is available, only a technically suitable alternative expressly agreed with the participant is used.

The practice agreement identifies the intended role and track, concrete learning objectives, tasks, agreed practice hours, supervision, permitted tools and data, confidentiality, approval boundaries, four milestones and required evidence.

A technically competent supervisor supports the practical work. Their suitability and responsibilities must be documented. The Academy supports milestone reviews and checks alignment with the learning goal. Supervising practice does not automatically authorise the supervisor to make the certification decision.

A weekly learning and work journal records tasks, time spent, methods, results, feedback and the next learning step. It excludes unnecessary personal or confidential information. The supervisor confirms traceable completion; attendance alone does not demonstrate competence.

By the end of practice week four

Milestone 1 – Scope and Baseline

Bounded assignment, baseline findings, learning objectives, data and approval status, and a traceable work plan.

By the end of practice week eight

Milestone 2 – First Substantive Output

Initial prototype, process design or governance output; recorded tests or reviews and justified adjustments.

By the end of practice week twelve

Milestone 3 – Validation and Improvement

Assessment against agreed criteria, stakeholder feedback, recognised limitations and documented improvements.

By the end of practice week sixteen

Milestone 4 – Handover and Reflection

Three completed track-specific outputs, a handover or continuation plan and personal reflection.

Evidence must identify the participant’s own contribution and disclose data sources, contributions by others and AI assistance. Confidential data is anonymised or replaced by suitable authorised evidence. For group work, individual responsibilities and assessment contributions are established in advance.

How your work is assessed

Weighting: 20%

A. Written Knowledge Component

Case-based short answers and reasoned decisions about AI use, output verification, data, roles and limitations.

45 minutes, supervised

  • Technically sound interpretation and clear reasoning.
  • Source and output verification and appropriate handling of uncertainty.
  • Critical: material data, approval or security boundaries in the case must not be ignored.

Weighting: 25%

B. Practical Specialist Task

A new bounded task suited to the selected track, with documented results and material working steps. The task bank uses equivalent variants and assessment criteria defined in advance.

60 minutes, supervised

  • Understanding of the assignment and a traceable approach.
  • Technical quality and verification of the result.
  • Critical: no unsupported claims or unauthorised production approval; necessary limitations must be stated.

Weighting: 35%

C. Practice Portfolio

Three track-specific outputs, four milestone records, the weekly journal and an overall reflection report. Business benefits must be supported by evidence or expressly identified as assumptions or scenarios.

Submitted in the agreed completion phase; the binding deadline is set before practice begins.

  • Clarity of objectives and suitability of the assignment (15% of the component score).
  • Technical execution and quality of the three project outputs (35%).
  • Traceable tests, reviews and improvements (25%).
  • Reflection, handover and transferability (25%).
  • Critical: data provenance, personal contribution and material limitations must be identifiable; practice hours or results must not be fabricated.

Weighting: 20%

D. Presentation and Professional Discussion

Eight minutes presenting the key results, followed by twelve minutes of technical questions including a changed scenario.

20 minutes

  • Clear communication suited to the audience.
  • Rationale for personal decisions and reflection on alternatives.
  • Critical: central working steps must be explained independently and transferred to a modestly changed task.

Pass requirements

A pass requires at least 70% overall after weighting and at least 60% in every mandatory assessment component. Rubric-based work uses a 0–4 scale: 0 = missing or technically unusable; 1 = major errors without a defensible rationale; 2 = substantially sound, with limited gaps; 3 = technically sound, traceable and applied appropriately; 4 = also critically evaluated, with justified alternatives and limitations. Every criterion expressly designated critical must score at least 2 out of 4. These pass thresholds are Academy programme rules; ISO/IEC 17024 does not prescribe them.

Permitted resources and AI

Written knowledge component: only expressly approved references; no generative AI or third-party communication. Practical task: AI and specialist tools specified in advance may be used where this fits the assessment objective; use and material inputs are recorded, and results must be checked independently. Portfolio: documented AI support and professional sources are allowed; personal contribution and responsibility must remain identifiable. Discussion: the candidate’s report and agreed presentation, with no undisclosed assistance.

Identity and independent work

Identity is checked before each supervised assessment component. Submitted work includes a declaration of authorship, a record of sources and tools, and identification of contributions by others. The professional discussion checks the candidate’s personal understanding of the submission. Personal and business data must be anonymised or submitted only with documented authorisation.

The common assessment rules also cover retakes, reasonable adjustments, appeals, confidentiality and certificate authorisation.

Assessment and Certificate Rules

What your certificate demonstrates

KI-Pass – Assessed Academy Certificate with Specialist Track

Scope of the credential
Records successful demonstration of common and track-specific competencies. It identifies the chosen track, teaching and practice hours actually completed, the type of practice environment, assessment date, programme ELA-PRG-KIP and revision. Project-based practice is not described as a workplace placement.
How to interpret the qualification
An Academy-owned assessed qualification. It is not a state vocational qualification, an IHK qualification or accredited personnel certification. Alignment with selected ISO/IEC 17024:2026 principles is not evidence of full conformity. External PECB certificates are separate credentials and require fulfilment of PECB’s own examination and certification requirements.
Assessment date and currency
Under Academy policy, this certificate is valid for a maximum of three years from its date of issue. It is issued only after all programme and assessment requirements have been met and the award decision has been documented. The validity period does not start on enrolment or at the beginning of the course. The actual validity period and expiry date are stated on the certificate; a shorter period applies only when expressly specified in the relevant course or assessment programme. There is no automatic extension. Renewal requires a documented review of current knowledge and application appropriate to the programme, using recent evidence and/or a suitable update or competence assessment. Repurchasing the entire course is not a blanket requirement. A renewed Academy certificate is also valid for no more than three years from its issue date. Attendance and passing an assessment remain verifiable historical facts tied to their dates; they do not provide indefinite validity as evidence of current competence. This rule applies only to the Academy-owned award; external certificates remain subject to their respective issuers’ conditions.

This programme applies to deliveries whose agreed course and assessment plan explicitly references this revision. Existing bookings retain their agreed terms.

Academy completion and external PECB credentials remain distinct

The foundation phase incorporates the PECB ISO/IEC 42001 Foundation and Lead Implementer learning routes with the teaching and preparation scope agreed before booking. Teaching, exam preparation, external examination and certification are different services. Whether PECB examination fees are included in the specific package is stated in the relevant offer. Passing the Academy KI-Pass does not replace a PECB examination; attending preparation does not automatically confer PECB certification.

View the current AI foundation package

The six-month standard model is offered individually with a chosen track, practice route, hours and complete price. The price and scope on the linked foundation-package page do not automatically apply to the complete qualification. Clarify the practice location, supervision, external examinations and any funding before booking.

Programme questions

Is this the same course as the current AI foundation package?

No. This standard model additionally includes specialist preparation, 16 weeks of supported practice and a completion phase. The existing foundation package is offered separately; inclusion or recognition of prior completion is expressly agreed before booking.

Which specialist track suits me?

Application Development concerns technical prototypes and integrations; Project Leadership concerns planning and delivery; Management and Process Design concerns workflows; Governance and Compliance concerns responsibilities, risks, approvals and evidence. The entry discussion considers your goal, existing knowledge and a suitable practical assignment.

Does the Academy guarantee a workplace placement?

No. A workplace placement is not automatically included. Suitable routes include an organisation, your own professional setting or an expressly agreed project environment approved for the learning objectives. The completion record identifies the practice actually undertaken.

Are 16 weeks of project work simply self-study?

Specialist practice is independent application with technical supervision, a weekly journal and four documented reviews. It is not an asynchronous teaching unit. Teaching and agreed learning sessions are live; practice and teaching hours are reported separately.

Can I use my own business or an existing project?

Yes, if the assignment, supervision, learning goals, data permissions and evidence are suitable and agreed before practice starts. Routine duties alone or attendance claimed retrospectively without verifiable work are not sufficient.

Do I automatically receive PECB certificates?

No. External examinations and certification requirements remain separate. Your offer states whether the examinations form part of the booked package. Academy completion and any PECB credentials obtained are recorded separately.

Is the KI-Pass an IHK or state vocational qualification?

No. It is an Academy-owned competence record with documented assessment and a specialist track. It does not assert IHK recognition, state equivalence or accreditation.

Can previous foundation learning be recognised?

A documented equivalence review is possible. Relevant factors are content, hours, currency and demonstrable competence. Recognition, remaining learning and assessment requirements are agreed in writing before booking; there is no automatic reduction.

Advice & course planning

Solo self-employed in Germany? Discuss potential KOMPASS support with an official advisory centre before booking. We can help you select a suitable programme in a consultation.

An Academy-owned assessed competence credential. The assessment programme refers to selected principles of ISO/IEC 17024:2026. This is not a claim of accreditation or fully demonstrated conformity.