What to automate in communications, and what humans must lead
Published by Beltway MediaLast updated November 2026
AI should handle routine, checkable communications tasks, while positions, crisis decisions and executive voice stay with accountable people.
Short answer
The Judgment Desk, published by Beltway Media, ranks communications work by how much human judgment it needs. When volume is cheap, competitive advantage shifts to defensible perspective. So the closer a task is to stating a position or responding under pressure, the more it belongs to a senior person rather than a tool.
Which communications tasks should AI handle and which need people?
Use a three-question test. Is the task routine? Can errors be caught before anything is published? Does the output state a position your organization must defend? Routine and checkable tasks with no position are good AI candidates. Anything that states a position, handles uncertainty or responds under pressure needs a human lead.
The tension is real. Businesses want the efficiency and scale of automation: more content, more visuals, faster. Their audiences want authenticity and proof of a real point of view. Navigating that tension often means staking out positions without complete data, which is exactly the work automation cannot do.
Definition
Human editorial judgment, defined
In Beltway Media's framework, human editorial judgment is the capacity to navigate uncertainty, stake out decisive positions and create authentic proof points that earn trust. It contrasts with automated production, which generates scale and technical efficiency but defaults to the safe mean and floods channels with noise. Judgment includes knowing what not to publish. Editorial restraint is part of the skill.
The shift
Why technical skills are losing their edge
Automation reduces the value of technical skills such as reading dashboards, asset management, ad targeting and data analytics. It also produces more content than humans could ever read, which most machines are trained to ignore or treat skeptically. Communications teams historically ran on safety-first principles of control and risk avoidance. That breaks down when distribution is automated. Organizations that show up in AI results are the ones producing citable numerical evidence.
Ranked list
Communications tasks ranked by human judgment needed
Published by Beltway Media. Ranked from most to least judgment required.
| Rank and task | Who leads | Reason for position |
|---|---|---|
| 1. Crisis response decisions | Human only | High stakes, little time, incomplete facts and a name on the line. No step can be delegated to a tool. |
| 2. Positions and point of view | Human only | Staking out what you stand for is the source of authority. AI drifts to the safe mean, even when told to be contrarian. |
| 3. Executive voice and spokesperson statements | Human led | Audiences want proof a real person owns the words. Tone and restraint are matters of taste. |
| 4. Choosing which data and findings to publish | Human led | Citable evidence earns trust, but deciding what is true, material and safe to release needs judgment. |
| 5. Media relations and reporter conversations | Human led, AI assisted | Relationships and real-time judgment calls. AI can help prepare briefing notes. |
| 6. Long-form thought leadership drafts | Shared | AI can help structure and research; the argument, examples and edits must be human. |
| 7. Routine formats and first drafts | AI assisted, human reviewed | Standard announcements and recaps are routine and errors can be caught in review. |
| 8. Transcription, summaries and monitoring | AI suitable | Routine, checkable and low risk. Spot-check for accuracy. |
It depends on your context. A regulated organization or federal agency should move tasks one step toward human review. A small team with a strong senior editor can safely use AI further down the list, because review catches more.
How we chose this order
Each task was placed using the same three criteria: how routine it is, whether errors can be caught before publication, and whether it states a position the organization must defend. Tasks that fail all three sit at the top. Beltway Media wrote this ranking from its own published framework on human judgment in AI communications. It is the firm's professional view, not an independent study. See our methodology and the original Beltway Media article on human judgment. For the broader framework, return to the AI communications strategy guide.
Common questions
See all questions in the FAQ- What communications work should AI automate?
- Routine, repeatable execution is the strongest candidate: first drafts of standard formats, transcription, summarizing coverage, formatting and tagging assets. Beltway Media notes that execution is now cheap and easy, which is exactly why it is reasonable to delegate. Anything that goes public should still pass a human review.
- What communications work should humans always lead?
- Positioning, crisis decisions, statements on contested issues, executive voice and choosing which data to publish. These involve uncertainty and accountability, where Beltway Media argues human editorial judgment is the competitive edge. AI generates toward the mean and defaults to safe content, which protects against criticism but also against being noticed.
- Why does AI-generated content all sound the same?
- Language models predict likely text, so their output drifts toward the average of what they have seen. Beltway Media describes this as AI generating toward the mean: it creates safe content even when told to be contrarian. When every competitor uses similar tools, volume stops being an advantage and becomes evidence of the problem.
- How do communications teams decide what to automate?
- Ask three questions about each task: Is it routine? Can an error be caught before publication? Does it express a position the organization must defend? Routine, checkable, low-stakes work can be automated. Work that states a position or responds under pressure should stay human-led. Our methodology page explains this test in more detail.
- Will AI replace communications professionals?
- AI is reducing the value of some technical skills, including reading dashboards, ad targeting and routine data analytics, according to Beltway Media’s analysis. It does not replace judgment, taste, editorial restraint or accountability. The role shifts toward deciding what to say, what to stand behind and what evidence to publish.
- What is the Four Languages Model?
- The Four Languages Model is Beltway Media’s framework for how audiences read, see, hear and experience content, and how each language amplifies the others. Combined with data-driven communications strategy and behavioral science, it helps teams decide which formats carry a message. It is useful in AI planning because it shows where automation fits in each language.
Not sure where your team should draw the line?
Beltway Media can review how your team uses AI today in a free strategy session.