M&A AI Governance · Director Liability · Transactions

The deal closes.
The liability doesn't.

During a merger or acquisition, AI is already running inside your transaction. The question is whether anyone is governing it — and whether you can prove it.

Understaffed teams under deadline pressure use whatever tools accelerate their work. In a transaction environment, that means sensitive financial data, confidential IP, and personal information flowing through consumer-grade AI tools with no audit trail, no data containment, and no governance architecture. Under Section 180 of the Corporations Act, the directors who sign off on that transaction hold the liability — regardless of which tool their team used, or whether they knew it was happening.

37%

AI Rework Rate

Of time saved through AI is offset by correcting its output — inside organisations using ungoverned AI tools daily.

Source: Workday, Beyond Productivity (Jan 2026, n=3,200)

$1.56M

Personal Exposure

Maximum personal liability for executives under the Financial Accountability Regime for ungoverned AI operations.

Source: ASIC · Financial Accountability Regime

Zero

Audit Trail

Consumer AI tools generate no court-admissible record of what the board saw, when, and what human oversight followed.

Oversight Obligation

01 The Problem Space

Transactions create the conditions for ungoverned AI to do its most damage.

The pressure is highest, the data is most sensitive, and the teams are most stretched. That combination does not reduce AI use — it accelerates it. What it eliminates is the governance layer that makes that use defensible.

Shadow AI enters the transaction

Due diligence teams, under deadline, use familiar consumer tools to parse legal contracts, summarise financial documents, and migrate data between systems. This is not reckless behaviour — it is normal human problem-solving under pressure. But in a transaction context, it creates immediate data privacy exposure and leaves no oversight record.

AI softens what the board needs to see clearly

When forensic findings are processed through an ungoverned AI tool, the tool's default behaviour produces neutral, palatable output. Material financial risks are reframed as areas for consideration. Liability indicators are smoothed into summary language. The board receives a version of the data — not the data. This is not a hypothetical risk: it is documented in our Blind Boardroom case study.

Stateful hallucination corrupts the merged entity

As AI sessions accumulate context across a complex transaction — hundreds of documents, multiple parties, overlapping data sets — the model begins to confuse earlier context with current inputs. This produces outputs that blend data from different sources, entities, or time periods. In a merger, that contaminated output can be embedded in the new corporate structure before anyone identifies the source.

The audit trail does not exist

When a regulator or opposing counsel asks what the board reviewed, when they reviewed it, and what human oversight accompanied that review — the answer, for organisations using ungoverned AI, is silence. Under Section 180, that silence is indistinguishable from wilful blindness.

02 The Regulatory Framework

The obligation is not emerging. It has arrived.

Directors and partners overseeing transactions that involve AI-assisted analysis carry a documented legal duty — in Australia under the Corporations Act, The standard is not intent. It is whether reasonable steps were taken, and whether those steps are provable.

Australia · Corporations Act

Section 180 — Duty of Care and Diligence

Directors are required to exercise their powers with the degree of care and diligence that a reasonable person would exercise. Where AI-generated analysis informs a board decision, the director holds the duty to ensure that analysis was accurate, complete, and subject to documented human review. "We used AI to process the report" is not a defence. It is a description of the governance gap itself.

Australia · Financial Accountability Regime

Personal Liability for Ungoverned AI Operations

Executives in the financial services sector face personal fines up to $1.565 million for operations — including AI-assisted operations — that lack adequate oversight architecture. Workslop is not only a productivity cost. In a regulated transaction context, it is a personal liability event.

Critical Distinction

A transcript is not governance.

Some organisations believe that saving AI chat logs constitutes an audit trail. It does not. A transcript proves the AI was used. It does not prove the output was verified, that the board reviewed the material findings, or that a human oversight event occurred. The regulatory standard requires documented oversight — not documented output.

03 The Vikings Solution

Governance architecture. Not policy. The mechanism.

Vikings of the Wire does not sell compliance checklists, AI software, or prompt libraries. We install the operational infrastructure that makes your organisation's AI use defensible — and measurably more productive. In a transaction context, that means three things delivered as a structured engagement.

Methodology

Fidelity Governance Protocol

Every AI-assisted analysis in your transaction is governed by protocols that preserve material data. Financial findings are not softened. Risk indicators are not reframed. The board receives the information as it exists — in the language of the finding, not the language of comfort.

Audit Trail

Fidelity Collaboration Score — Evidentiary Record

Every AI-assisted analysis generates a timestamped, documented record of the oversight event — what was reviewed, when, and what human verification accompanied it. This is produced automatically as the governance events occur. It is not retrospective reconstruction. It is real-time proof.

Capability

Fidelity Collaboration Steward

We train a designated person within the transaction team — or the acquiring organisation — to operate and maintain the governance architecture from the inside. External dependency reduces. Internal capability compounds. The organisation owns its own protection after we leave.

04 Data Integrity — Available Capability

For transactions where data containment is non-negotiable.

Where a transaction involves data that cannot leave the client's infrastructure — sensitive PII, confidential IP, regulated financial records — Vikings has built and operates a stateless pipeline architecture that processes data without retaining it. This is available capability for engagements that require it, not a standard offering.

How It Works

Stateless Pipeline — Client Infrastructure, Zero Residual Footprint

Rather than sending data to an external service, we connect our secure AI pipeline directly to your existing infrastructure — AWS S3, Azure, Google Workspace, or SharePoint. We request access to a designated folder only. Data is processed through enterprise-grade AI APIs where model training on client data is contractually disabled. Session memory is cleared after each processing node executes. Zero chat history. Zero context accumulation. Zero residual data footprint. For organisations already operating on Google Workspace, the data containment guarantees are embedded in the enterprise agreement by default.

Positioning Note

This capability exists. It is not the primary proposition.

The primary value Vikings delivers in a transaction context is governance methodology and audit trail infrastructure — not data pipeline technology. The stateless pipeline is available for engagements where data containment is a specific client requirement. Most organisations do not need it. All organisations need the governance layer.

05 The Commercial Case

This governance layer pays for itself before it protects you.

The rework cost of ungoverned AI is measurable, consistent, and significantly larger than the cost of governing it. In a transaction team operating under deadline pressure, that calculation is straightforward.

The Rework Tax — Illustrated

Workday's January 2026 study of 3,200 knowledge workers found that 37% of time saved through AI is offset by rework — correcting, verifying, and reconstructing AI output that was not governed at the point of production. For a transaction team of 20 professionals at $300/hour average billing rate, working a compressed 12-week due diligence cycle, that rework tax is not an abstraction. It is a calculable line item that arrives before the deal closes — and before any regulatory exposure is triggered.

1.5

Weeks Lost Per Worker

Annual rework burden per highly engaged AI user. In a compressed transaction timeline, this materialises within weeks.

20%

Vikings Fee Model

We charge 20% of recovered workslop. No results, no fee. The governance layer is self-funding from rework reduction alone.

9.2

Average FCP Score

Average Fidelity Collaboration Score at trial completion — the measurable output quality benchmark our methodology delivers.

06 For Capital Partners

Mandate governance as a condition of capital.

Investment banks and venture capital partners deploying capital into organisations that use AI operationally carry indirect exposure to the governance gaps in their portfolio companies. The organisations most likely to face regulatory action under Section 180 are those operating AI without the architecture to prove oversight. That risk does not stay inside the portfolio company.

The Mandate Model

Governance as a deal condition

Vikings governance assessment and implementation can be structured as a condition of capital deployment — either as a pre-condition of closing or as a 90-day post-close requirement. This gives the capital partner documented proof that the portfolio company's AI operations are governed, auditable, and defensible. Not a vendor contract. Not a general AI policy. A paper trail.

What You Receive

Documented governance architecture

A structured governance assessment identifying current AI exposure across the organisation. A clear roadmap to close the gaps. Implementation of the Fidelity Governance Protocol and audit trail infrastructure. Training of the Fidelity Collaboration Steward. A governance report suitable for board-level review and regulatory response. The initial assessment carries no obligation.

The deal closes.The liability doesn't.