Recommended Free Tools
Giving an AI “morals” is not simply a matter of writing down rules and teaching a model to follow them. It requires deciding what the system should align with—instructions, people’s preferences, their interests, or broader values—and whose judgments should shape those choices. A scholarly framework helps clarify that problem, but it does not establish what Anthropic has done: Iason Gabriel’s 2020 article examines AI alignment generally, not Anthropic’s methods or decisions.
What does it mean to give an AI moral guidance?
“AI alignment” can describe several different goals. A system might be expected to follow explicit instructions, carry out a person’s intentions, reflect preferences people have expressed, serve their interests, or act according to values. These goals can point in different directions. A request may conflict with someone’s longer-term interests, for example, while a preference expressed by one person may not represent the views of others affected by the system.
As an Amazon Associate I earn from qualifying purchases.
That makes alignment partly a normative question: deciding what a system ought to do, and whose judgment should count, cannot be settled by technical implementation alone. Gabriel’s 2020 analysis argues that the aim should not be to discover and encode one uncontested moral theory. Instead, the challenge is to identify fair principles that people could reflectively endorse despite disagreement about morality. He describes the problem as political rather than metaphysical: it concerns legitimate ways to make shared rules, not just the discovery of a single moral truth. Gabriel, “Artificial Intelligence, Values, and Alignment,” Minds and Machines, published 1 October 2020.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsWho should decide which principles an AI follows?
Gabriel discusses three broad approaches. Each handles disagreement differently, and none removes the need to make choices about inclusion, fairness, and legitimacy.
#1 Best Overall
Look for principles with broad support
A public morality grounded in human rights seeks principles that can command support across different moral and philosophical traditions. Its strength is the possibility of shared protections without requiring everyone to agree on a complete moral doctrine. The hard question is which principles genuinely have broad support and how to interpret them when people disagree about their application.
Aggregate people’s moral judgments
Another option is to collect individual judgments and combine them. That immediately raises practical and ethical questions: Whose judgments are included? How are they elicited and measured? What method combines them? An aggregate result can look like a collective answer while concealing whose views were excluded or given greater influence.
Rank #2
Use democratic procedures
Voting, public discussion, and civic engagement can give people a role in developing AI principles. But a procedure is not automatically legitimate merely because it is democratic in form. Its fairness, participation, and the degree of endorsement required for foundational rules all matter. Any process also needs to address how principles remain credible as circumstances and public views change.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →How can these approaches be evaluated?
A useful comparison asks what each approach makes possible—and what remains unsettled—rather than treating any one method as a complete solution.
| Approach | Who contributes? | Key challenge | What it may offer |
|---|---|---|---|
| Broad public morality and human rights | People and traditions whose support can be reflected in shared principles | Determining what support is genuinely broad and how to apply principles amid disagreement | A common foundation without requiring agreement on one full moral theory |
| Aggregation of moral judgments | Individuals whose judgments are collected | Deciding who is included, how judgments are measured, and how they are combined | A way to represent multiple judgments, provided the choices behind the aggregation are made explicit |
| Democratic procedures | Participants in voting, discussion, and civic engagement | Establishing a fair and legitimate process and deciding how much endorsement foundational rules need | A process through which people can help develop and contest principles |
Across these options, Gabriel identifies qualities a principle-selection process should pursue: procedural fairness, genuine inclusion, concrete guidance, stability, robustness, and comprehensiveness. These goals can pull against one another. A broadly inclusive process may be difficult to organize; a concise rule may be easier to apply but fail to address important cases. The framework therefore directs attention not only to the resulting principles, but also to how they were selected and whether affected people can endorse the process.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What this framework does—and does not—say about Anthropic
Gabriel’s article is scholarly context for thinking about AI values and alignment. It does not report on Anthropic, identify the company’s initiatives, describe its models’ behavior, or verify statements by its staff. It therefore cannot establish that Anthropic has adopted any of these approaches or show how the company gives its AI moral guidance. The title’s reference to Anthropic should not be read as evidence of a particular company practice.
What the framework does clarify is why the phrase “give AI morals” can mislead. There is no self-defining list of human values waiting to be converted into code. The work involves choosing an alignment target and developing principles through a process that can address disagreement fairly. Those choices are as consequential as the technical systems that implement them.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
