What brings you here?
For citizens
How can citizens understand what happens between a leader’s promise and an outcome for residents? ObliNet presents an approach connecting delegated authority, decisions, commitments and evidence of fulfillment. AI is proposed as a tool for analyzing decision grounds and consequences. The aim is to make the use of public authority easier to understand and assess, while protecting personal data and allowing erroneous conclusions to be challenged.
For leaders
How can leaders make well-founded decisions when information is incomplete, departments understand the situation differently and execution depends on many participants? ObliNet proposes an approach to multilevel management: check decision grounds, authority, resources and dependencies before confirming commitments, then track changes and outcomes. The site explores an AI assistant that compares options and reveals possible consequences. Leaders retain choice and responsibility, while the grounds for their decisions become available for scrutiny.
For politicians and elected representatives
How can a political program become feasible commitments and outcomes that citizens can verify?
ObliNet proposes supporting decisions from situation analysis and checks of authority, resources and dependencies through execution and assessment of consequences. AI helps compare options, identify contradictions and prepare explanations; authorized people and institutions retain choice and responsibility.
Accountability is treated as an ongoing part of political work: the public should be able to understand what a leader promised, the grounds for action, what changed and what was achieved. This allows achievements and mistakes to be discussed in the context of the decision. The approach’s practical effectiveness remains to be tested in pilots.
For initiatives to improve elections
How can the choice of representatives become better informed, with a continuing connection between voting and elected representatives’ accountability?
ObliNet proposes considering elections alongside subsequent scrutiny of delegated authority. Verifiable commitments, decision grounds and outcomes could help citizens compare programs with actions, distinguish broken promises from justified revisions and ask substantive questions of candidates and serving representatives.
AI may help analyze evidence, while assessments remain open to scrutiny and challenge. This is a basis for investigating better-informed elections and accountability between them. It does not offer a ready-made electoral reform or a single algorithmic rating of politicians: procedures for electing, overseeing and replacing representatives require their own public deliberation.
A continuous cycle
In public governance, ongoing assessment of decisions and outcomes complements periodic elections. Existing laws and institutions retain their roles.
- Citizen mandate
- Delegated Authority
- Explicit Commitments
- Observable decisions
- Execution
- Outcomes
- Evidence
- Public Confidence
What does “improve healthcare” mean?
What exactly will improve? By when? Against which criteria? Which resources and permissions are required? What is within the leader’s control, and what depends on others? You cannot accurately assess fulfillment of an obligation that was never clearly defined.
Before, at and after a decision
Before: preserve a Decision-Time Snapshot containing the problem, available Evidence, unknowns, alternatives, assumptions, risks and Dependencies. At the decision: distinguish AI advice, expert advice, the leader’s choice and its reasoning. After: track Execution, Outcomes, forecast differences and learning.
Citizen mandate delegates Authority. A Decision-Time Snapshot informs the human or institutional Decision. AI analysis supplies a recommendation from the side. Execution produces Evidence that informs, but does not guarantee, Public Confidence.
In ObliNet, the commitment is the primary object of coordination. Its feasibility is checked before confirmation; afterward, dependencies, execution, changing conditions and confirmation of the result are tracked.
AI helps participants formulate, check and align commitments, and assess evidence of their fulfillment.
The authority to enter into or confirm commitments, and to accept fulfillment, remains with the relevant authorized participants.
An objective assistant for leaders
Public and government leaders need an assistant that helps scrutinize decisions, including uncomfortable alternatives and consequences for different groups of citizens. The interests of a leader, an agency, a technology provider and the public may differ. Objectivity is therefore a requirement for a checkable process, not an automatic property of AI.
ObliNet proposes linking recommendations to sources, assumptions, authority and commitments at every level of management. Resources, conditions and dependencies are checked before a decision; affected commitments are reconsidered as circumstances change. Alternatives, reasons for choosing and evidence of outcomes are retained. Facts, forecasts and value priorities must remain distinguishable.
Independent review, disclosure of conflicts of interest, opportunities to challenge conclusions and the ability to replace a model or provider are needed. Comparing models may help, but agreement does not prove correctness. Access to data and decision histories must respect privacy and legitimate restrictions. Authority and responsibility remain with the appropriate people and institutions.
AI should reduce routine preparation of records and approvals. The depth of checks should match the cost of error and the reversibility of a decision; effectiveness is assessed through the combined losses from errors, delays and verification costs.
Evaluate the decision, not only the outcome
A reasonable decision can have a bad outcome. Accountability should examine what was known at the time, the alternatives considered and the treatment of uncertainty. A successful outcome alone does not excuse negligence.
Leadership as public service
Public work rests on temporarily delegated Authority. Greater delegated Authority means greater responsibility and stronger obligations of observability and explanation. The aim is to make governance closer to a profession and vocation, with less room for private advantage. Whether this changes who seeks office must be tested.
Room to be wrong — and to learn
Reasonable error, lack of competence, negligence and deliberate deception are different. “We relied on assumption X. New evidence contradicted it, so we changed course.” Detecting and correcting an error should count as a strength.
Public Confidence, without a single score
Public Confidence is multidimensional and informed by evidence; it is not an algorithmic political score. Facts and values are different: citizens may agree on what happened and disagree about whether it was desirable. AI must never automatically remove a leader from office.
Delegated Observation
Citizens may delegate analytical attention to experts, institutions or AI services; Delegated Observation must remain revocable. An Observer analyzes without acquiring decision Authority. An Auditor has a formal Mandate to examine defined Evidence or processes. Observers must disclose methods, conflicts and Evidence and remain open to audit.
Public power. Private lives.
Transparency of public power must not become transparency of private citizens. Use proportionate disclosure, access boundaries and opportunities to challenge errors. Public Confidence Review remains a human and institutional process.