How your AI is governed is being assessed whether you present it or not – by customers, by investors, by acquirers. We help teams building AI, and those backing them, put one answer in front of all of them.
Request a fit checkWe work hands-on with a small number of clients. After a short call, you will hear a clear yes or no.
Is this high-risk under our own policy?
What is the exposure, and what will it cost?
Will it hold up after closing?
Which tier, and on what basis?
One answer, from one evidence base.
You build and sell it.
Walk into procurement, diligence, and board conversations with one posture, not four answers.
Added to what you already sell.
Answer tender and procurement questions from one conclusive evidence base: the same story in every answer, and each answer quick to produce instead of rebuilt from scratch.
You invest in AI companies.
Assess a target’s exposure before signing; map it across the portfolio; report to LPs with substance.
You incubate or accelerate teams.
Get cohort companies through procurement and diligence sooner, with a posture they keep after the programme.
We understand that you might sit on both sides. We scope each engagement to the side you prioritise at this moment.
Why
Buyers, investors, and acquirers increasingly ask not whether you take AI governance seriously, but what you can show them. Regulation such as the EU AI Act sets legal requirements. Procurers, investors, and acquirers set their own, and increasingly check them before they commit.
It tells you where you stand against the rules. It does not tell you what to put in front of a buyer, an investor, or an acquirer when they ask.
A memo for the regulator, a slide for the investors, a questionnaire for the buyer. The answers drift, confuse your own team, and the gap shows up in diligence.
What moves a decision is what can be shown: risk tier, human oversight, training data, and what happens when something goes wrong.
What we do
We co-create a single posture with you: what you can demonstrate today, to whom, where the gaps are, and what to close first. From there we answer buyers, investors, regulators, and users out of the same evidence base.
The phasing is the same on both sides; the cadence differs. Builders engage across product cycles. Backers engage transaction by transaction, or at fund-wide checkpoints.
Align decision owners, constraints, and your expectations for your product or portfolio.
Set the posture: what you can demonstrate today, to whom, where the gap is, and what to close first.
Close the priority gaps and put the evidence into product operations.
Keep the posture current as the product, the portfolio, and the expectations move.
| Evidence | The buyer asks | The investor asks | The acquirer asks | The regulator asks | What you show | EU AI Act |
|---|---|---|---|---|---|---|
| Risk tier | Is this high-risk under our own policy? | How exposed is the company, and what will compliance cost? | Will the classification hold up after closing? | Which tier, and on what basis? | Your classification, with the assumptions stated | Art. 6, Annex III |
| Human oversight | Who signs off on what it outputs? | Is oversight built in, or bolted on? | Does oversight depend on people who might leave? | Can a person intervene, and is it logged? | Who can intervene, when, and how it is logged | Arts. 12, 14, 26(2) |
| Training data | Is our data used to train your model? | Are the data rights clean enough to scale? | Do we inherit any data liabilities? | Where does the data come from, and is it documented? | Sources, rights, and known gaps | Art. 10; GDPR |
| When something goes wrong | What happens if it fails inside our process? | What would an incident cost? | Is there a record of past incidents? | How are serious incidents reported? | Monitoring, escalation, and the incident record | Art. 72–73 |
The question changes with the room. What you show does not.
How we work
Fixed-scope sprints, workshops, and advisory. Each one ends in a decision you can take, not a report for the shelf.
Both founders deliver every engagement. No junior layer.
EU digital policy expertise alongside operating experience in product, sales, and due diligence.
Not a law firm. Where legal counsel is needed, we say so and work alongside it.
Builder side, sprint
A fixed-scope assessment that sets your posture: what you can demonstrate today, to whom, where the gap is, and what to close first.
ServesBuyerInvestorRegulator
Example: a page from a readiness assessment.
Builder side, workshop
Working sessions that put the posture into product operations, so the evidence is produced as the product is built rather than assembled when someone asks.
ServesEvery room
Example: a team enablement session.
Backer side, sprint
A focused EU AI Act exposure read on a single target. Surfaces deal-breakers, conditions, and post-investment risk before signing.
ServesInvestor
Example: a diligence memo.
Backer side, retainer
Portfolio risk you can price. Ongoing mapping that makes governance legible across a pipeline or portfolio, and feeds what you report upward.
ServesInvestorAcquirer
Example: a portfolio clinic.
Who we are
We built AIffect because we kept seeing the same pattern on both sides of the table: AI teams without a good answer to what they can demonstrate, and AI investors without a good way to assess it.
Founding Partner
Turns AI regulation into
product-level decisions.
Founding Partner
Translates between technical realities and strategic decisions on AI and digital governance.
If you build AI, embed it in what you sell, or back teams that do – probably. Many companies do both; we scope each engagement to the side you are working on. If you sit elsewhere on the value chain, send a brief and we will tell you quickly.
Yes. The EU AI Act can apply to companies outside the EU when they sell into the EU market. Where it does not, the question usually arrives anyway: if you sell to European customers, raise from European investors, or expect a European acquirer, their questions will be shaped by it. We work from the EU AI Act as the benchmark, and alongside local counsel where other jurisdictions matter.
A legal opinion tells you where you stand against the rules. We tell you what to put in front of a buyer, an investor, or an acquirer when they ask, and we build it so the same answer holds in every room. We work alongside legal counsel where it is needed, not instead of it.
Compliance is part of it. But regulation – the EU AI Act being the most developed example – is one set of requirements among several: enterprise customers, investors, and acquirers each have their own, and they often check them earlier and in more detail. We work on all of them together, with regulation as the baseline.
Yes. We run portfolio clinics that map EU AI Act exposure across your active investments, flag urgent remediation cases, and feed your limited partner and investment committee reporting. We can also support single-target pre-investment diligence on a transaction timeline.
Technical due diligence validates the product. AI Act due diligence validates the deal – whether the target’s risk classification, evidence trail, and governance maturity will survive enterprise procurement, acquirer scrutiny, and supervisory attention. Different question, different output.
Yes: judgement first, AI-augmented. We use AI to work faster and more thoroughly – to scan regulation and standards, structure evidence, and draft and cross-check documents. Every conclusion, recommendation, and document you receive is reviewed and owned by one of us. We do not put client material into AI tools that train on it.
We are based in Berlin and work remotely with clients across Europe and internationally. Where it makes sense, we work on-site with you and your team – particularly for workshops and decision-intensive sessions.
If you build AI, embed it, or back the teams that do, request a fit check.
Request a fit check