TAER

Ethically Sourced™ Safe AI in Football. Part of An IBM Partner Programme

Ethically Sourced™ AI TAER Watch in Football

As AI Agents enter football operations, recruitment, commercial activity and fan engagement, who keeps watch? What Stops Rogue AI Activity?

Team Agent Emergency RevocationFootball AI SupervisionHuman AccountabilityFootball Operations

Why Football Needs an AI Control Layer

AI Agents are moving beyond answering questions. In football, they could increasingly support recruitment and scouting, player and match analysis, ticketing, sponsorship, procurement, communications, fan engagement, academy operations, travel, commercial negotiations and other club or governing-body activities.

This creates a new operational challenge for clubs, leagues, associations and other football organisations: What happens when an AI Agent does something nobody intended, nobody authorised—or nobody notices until it is too late?

Beyond “Rogue AI” in Football

Recent AI security testing has shown why increasingly capable Agents require strong boundaries, monitoring and intervention controls. In football, the immediate issue is not science-fiction “rogue AI”; it is whether an Agent acting for a club or football organisation has been given excessive access, insufficient supervision or authority beyond what was intended. When an Agent acts unexpectedly, we should ask:

  • Authority: Which club, governing body or other football organisation gave it authority?
  • Access: Could it access player data, scouting systems, commercial systems, communications, ticketing, finance or other sensitive resources?
  • Boundaries: What sporting, commercial, regulatory and technical boundaries were in place?
  • Monitoring: Who or what was monitoring its activity?
  • Intervention: Who could stop it quickly if necessary?

A Fundamental Shift in Governance

A football AI Agent operates with objectives, tools, credentials, permissions and access provided by people and organisations. One Agent might analyse recruitment targets, another support sponsorship, another communicate with supporters, another assist procurement, and another negotiate with authorised counterparties or other AI Agents.

The core governance question is no longer simply: “Did the AI give the right answer?”
It has become: “Is this AI Agent acting within the authority this football organisation intended to give it?”

Football Cannot Rely on Humans Watching Every Agent Action

Good AI governance in football starts with preventative controls: verified identities, least-privilege access, defined authority, containment and clear human approval thresholds. But safeguards can fail. Permissions can be misconfigured, software can contain vulnerabilities, Agents can find unexpected routes to an objective, and multiple Agents can interact in unforeseen ways at machine speed.

AI-Scale Activity Requires AI-Scale Supervision

AI Agents can perform thousands of actions over short periods. As football organisations deploy Agents across sporting, administrative and commercial functions, the volume and speed of activity could quickly exceed what staff can supervise transaction by transaction.

At that scale, simply saying that a human is “in the loop” is not enough. Humans should remain accountable, but they cannot realistically observe every action of every Agent around the clock. AI-scale football operations require AI-scale supervision.

Introducing the Ethically Sourced™ AI TAER Agent

TAER introduces a dedicated supervisory capability into a football organisation’s AI team. The Ethically Sourced™ AI TAER Agent is an AI Agent whose job is to help watch over other AI Agents. It is a control-plane Agent rather than a player-recruitment, ticketing, sponsorship, procurement or other operational Agent.

WHAT AN AGENT IS AUTHORISED TO DO <── VS ──> WHAT THE AGENT IS ACTUALLY DOING

Specific Warning Signs & Detection Triggers

The TAER Agent can monitor for football-specific operational anomalies such as:

  • Authority Breach: Is a recruitment, procurement, commercial or negotiating Agent exceeding its delegated authority?
  • Unauthorised System Access: Is it attempting to access player, scouting, medical, financial, ticketing or other systems outside its permissions?
  • Bypassing Approvals: Is it attempting to bypass a sporting director, club executive, safeguarding, legal, finance or other required human approval?
  • Mission Drift: Is an Agent commissioned for one football purpose drifting into another without authority?
  • Counterparty Risk: Is it interacting with an Agent, intermediary or digital counterparty whose authority or Safe AI Passport has been suspended?
  • Velocity Anomaly: Has its volume of messages, offers, enquiries, transactions or system actions changed unexpectedly?
  • Privilege Escalation: Is it attempting to obtain additional club permissions, credentials or data access?
  • Multi-Agent Anomaly: Are several club or counterparty Agents exhibiting correlated abnormal behaviour?

Watch. Detect. Intervene.

TAER is designed around a straightforward operational sequence that can help a club or football organisation retain control:

1

WATCH

Observe significant Agent activity, system use and Agent-to-Agent interactions across authorised football operations.

2

DETECT

Identify behaviour that falls outside the Agent’s authority, football organisation policy or expected operating parameters.

3

ALERT & HOLD

Notify the accountable club or football official while pausing a questionable action, communication or transaction.

4

RESTRICT & ISOLATE

Remove permissions or capabilities and separate the Agent from club systems, tools, data or other Agents.

5

REVOKE, PRESERVE & RECOVER

Withdraw delegated authority, preserve relevant evidence and enable controlled investigation, review and recovery.

Passport Integration & Independent Enforcement

To recognise abnormal behaviour, a TAER Agent needs to know what normal, authorised behaviour looks like. The Ethically Sourced™ Safe AI Passport can provide that baseline by identifying the Agent, its accountable football organisation, its purpose, delegated authority, restrictions, approval requirements and current TAER status.

IDENTITY ──► AUTHORITY ──► ACTIVITY ──► ASSURANCE

Independently Enforceable Controls

If a football AI Agent behaves unexpectedly, emergency control should not depend on that Agent agreeing to shut itself down. TAER should connect to independently enforceable controls capable of withdrawing:

Credentials
Permissions
Tool Access
Transaction Authority
Agent-to-Agent Comms
Mission Scope

Controlling the TAER Agent Itself

A TAER Agent capable of restricting other football Agents is itself powerful and should carry its own Ethically Sourced™ Safe AI Passport with strictly bounded authority. It may be authorised to monitor, alert, hold, isolate and preserve evidence, but should not be able to grant itself new powers, increase its own authority, erase evidence or rewrite audit history.

“A TAER Agent should be powerful enough to remove authority—but not powerful enough to grant itself authority.”

The DOER • CHECKER • STOPPER Model

TAER introduces a practical framework for structuring safer multi-Agent teams across football operations:

Role Entity Operational Function
DOER Operational AI Agent Performs authorised football, administrative or commercial tasks.
CHECKER Validation Agent / Assurance Checks important outputs, delegated authority and club or governing-body policy compliance.
STOPPER TAER Agent Independently watches for conditions requiring intervention, restriction or emergency revocation.
HUMAN Accountable Owner Retains ultimate organisational authority and accountability.

Keeping Humans Meaningfully in Control

As football organisations deploy more AI Agents, the challenge is keeping their identities and authority identifiable, bounded, observable and revocable. TAER does not remove human accountability—it is intended to help make human accountability practical at AI speed.

IDENTIFY ──► VERIFY ──► DEFINE ──► WATCH ──► DETECT ──► ALERT ──► HOLD ──► ISOLATE ──► REVOKE ──► PRESERVE ──► KEEP HUMANS IN CONTROL

Don’t just trust an AI Agent in football.
Verify it. Monitor it. Be able to stop it.

Ethically Sourced™ Safe AI for Football • Identifiable • Accountable • Evidence-Backed • Monitored • Revocable