A government agency running a social media account has to write its policy against several overlapping bodies of law: the First Amendment (which governs comment moderation and blocking), the Supreme Court’s state-action test from Lindke v. Freed, the Hatch Act for federal employees, federal and state public records laws, Section 508 and ADA Title II accessibility standards, and the Privacy Act. Each of these creates specific rules a written policy has to address, and each carries its own penalties when it’s ignored. The legal requirements for a government social media policy are less about picking a platform and more about drawing lines that hold up when someone sues.
Comment Moderation and the First Amendment
When an agency turns on comments, the comment section becomes what courts treat as a designated public forum. The controlling rule there is straightforward: the government cannot discriminate based on viewpoint. A city page cannot delete comments criticizing the mayor while leaving supportive ones up.1Congress.gov. Constitution Annotated – Amdt1.7.4.1 Overview of Viewpoint-Based Regulation of Speech Even speech on topics the agency would rather avoid is protected when it reflects a perspective the commenter actually holds.2Congress.gov. Constitution Annotated – Amdt1.7.7.2 Public and Nonpublic Forums
Moderation is still possible. It just has to be content-neutral, written down in advance, and applied to everyone the same way. Removing commercial spam, off-topic solicitations, obscenity, and genuine threats of violence is defensible. A policy that lets administrators delete anything they consider “inappropriate” without defining the term is not.
What Counts as a Removable Threat
True threats can always be removed, but the legal definition is narrow. In Counterman v. Colorado (2023), the Supreme Court held that a true threat requires the speaker to have consciously disregarded a substantial risk that the communication would be seen as threatening violence.3Supreme Court of the United States. Counterman v. Colorado, No. 22-138 Angry criticism does not qualify. “The entire council should be fired” is protected. A specific, credible threat of physical harm is not. When in doubt, leaving the comment up and consulting counsel is safer than deleting it, because removing protected speech carries far greater legal risk than tolerating an unpleasant post.
When an Account Is Actually “The Government”
Not every account run by a public employee is subject to the First Amendment. The Supreme Court set out a two-part test in Lindke v. Freed (2024), and both parts have to be met before an official’s social media conduct becomes state action.
First, the official has to have actual authority to speak on the government’s behalf about the subject of the post. What the employee was actually hired to do controls, not what the job description theoretically allows. Second, the official has to be exercising that authority in the particular post at issue. If not, the Court wrote, “he speaks with his own voice.” Announcing something available only through the agency looks official; sharing publicly available news does not, even from the same page.4Supreme Court of the United States. Lindke v. Freed, No. 22-611
For policy drafting, this pushes hard toward separation. An account that passes to the next person in the job is almost certainly government action. An account that mixes family photos with official announcements creates the ambiguity that ends up in court. The cleanest approach is to keep official and personal accounts entirely separate and to name in the policy exactly which accounts are authorized to speak for the agency.
Turning Comments Off as a Policy Choice
When an agency speaks for itself instead of hosting a forum, the First Amendment’s forum rules do not apply. The Supreme Court confirmed this in Pleasant Grove City v. Summum.5Justia Law. Pleasant Grove City v. Summum, 555 U.S. 460 An agency can run an account with comments disabled and post whatever it chooses. The moment comments are turned on, forum rules kick in and viewpoint-neutral moderation becomes mandatory. Agencies that want the control that government speech allows should make that choice explicitly in the policy and configure their pages to match.
Employee Speech on Personal Accounts
Public employees don’t lose their First Amendment rights by taking the job, and a social media policy that tries to strip them does not survive challenge. Courts apply the Pickering-Connick balancing test: speech on a matter of public concern is protected, weighed against the agency’s interest in workplace efficiency.6Congress.gov. Constitution Annotated – Pickering Balancing Test for Government Employee Speech The more the speech addresses a real public issue rather than an internal workplace grievance, the harder discipline is to justify.
That protection disappears when the employee is speaking as part of the job. Under Garcetti v. Ceballos, statements made in the course of performing official duties get no First Amendment protection at all, even on matters of public concern.7Justia Law. Garcetti v. Ceballos, 547 U.S. 410
A social media policy can require a disclaimer on personal accounts that opinions are the employee’s own. It can prohibit disclosure of confidential job information. It can restrict conduct that genuinely impairs agency operations. What it cannot do is broadly forbid employees from criticizing the agency or its leadership on their personal accounts when they speak as private citizens on matters of public interest. Policies that reach further get struck down.
Hatch Act Rules for Federal Employees
Federal employees have to follow the Hatch Act on top of everything else. The statute prohibits most federal employees from using their official authority or influence to affect an election, and from engaging in political activity while on duty, in a government building, or on government equipment.8Office of the Law Revision Counsel. 5 USC 7323 – Political Activity Authorized; Prohibitions
On social media, that means no sharing a campaign fundraiser from a government-issued phone, no partisan endorsement posted from a government computer, and no using an official title to promote a candidate on any platform. Employees at certain agencies, including the FBI, CIA, and Secret Service, face stricter rules barring them from active participation in political campaigns.
Penalties are real. They range from reprimand to removal and a ban on federal service for up to five years, and a civil penalty of up to $1,000 can be added.9Office of the Law Revision Counsel. 5 USC 7326 – Penalties A policy that says only “follow the Hatch Act” is not doing its job. It should spell out that posting from a personal device is not automatically safe, because the critical factors are whether the employee is on duty or in a government space and whether they are leveraging their official position.
Social Media Posts Are Public Records
Content an agency creates or receives while conducting public business is generally an official record regardless of format. The Freedom of Information Act gives the public a right to request records from federal agencies, and every state has a comparable transparency law.10FOIA.gov. Freedom of Information Act Frequently Asked Questions Posts, replies, direct messages, and account metadata all fall within this framework when they relate to government business.
The hard part is preservation. Platforms are not records-management systems. Posts can be deleted, accounts suspended, and platform changes can erase years of content overnight. A policy should require dedicated archiving tools that capture content in real time and preserve it in a searchable format. Retention depends on the content’s function rather than its format. A post announcing a public hearing may need to be kept as long as any other public notice, and retention schedules vary by jurisdiction. Failure to preserve and produce records on lawful request can lead to court orders and financial penalties, the same exposure that comes with destroying paper files.
Accessibility Requirements That Apply Now
Government social media has to be accessible to people with disabilities. Section 508 of the Rehabilitation Act requires that any information and communication technology created by federal agencies be usable by people with disabilities.11Section508.gov. IT Accessibility Laws and Policies That covers everything posted to social media.
State and local governments have a parallel duty under Title II of the ADA. A 2024 Department of Justice rule established WCAG 2.1 Level AA as the binding technical standard for government digital content, including social media. Governments serving 50,000 or more people must comply by April 24, 2026. Smaller governments and special districts have until April 26, 2027.12ADA.gov. Fact Sheet – New Rule on the Accessibility of Web Content and Mobile Apps
Federal guidance from Section508.gov identifies specific requirements a social media policy needs to build in:
- Every image or GIF needs concise alt text that conveys the image’s meaning to screen reader users without repeating what’s already in the post.
- Videos need synchronized captions, either closed or open, and audio description when the visual carries information the narration does not.
- Each word in a multi-word hashtag has to be capitalized (e.g., #AccessibleGovernment) so screen readers parse the words.
- Text in images has to meet a minimum 4.5:1 contrast ratio for regular text and 3:1 for large text.
- Strings of emojis become an unintelligible stream through a screen reader, and GIFs that flash can trigger seizures with no user control to pause them; both need limits.
These are binding legal requirements. Agencies that ignore them risk disability discrimination complaints.13Section508.gov. Social Media
Privacy Act Limits on Collecting Information
Federal agencies that collect information about individuals through social media have to comply with the Privacy Act of 1974. The law limits agencies to maintaining information about individuals that is relevant and necessary to a purpose authorized by statute or executive order. When an agency collects personal information, including through monitoring or direct messages, it has to tell the individual the legal authority behind the collection, the purpose, how the information will be used, and the consequences of not providing it. Agencies also have to publish notices of their records systems in the Federal Register.14Office of the Law Revision Counsel. 5 USC 552a – Records Maintained on Individuals
The practical trap: an agency that monitors public comments and compiles the information into files organized by individual name or identifier has created a “system of records” under the Act, with all the obligations that come with one. A social media policy should say clearly what information staff may collect from social media interactions, how it’s stored, and who can access it. Casual collection that looked harmless becomes a legal obligation once it becomes systematic.
What Happens When the Policy Fails
When a government official blocks a user or deletes a comment in a way that qualifies as state action under Lindke, the person affected can file a federal civil rights lawsuit under 42 U.S.C. ยง 1983.4Supreme Court of the United States. Lindke v. Freed, No. 22-611 These suits can seek both injunctive relief (an order to unblock or restore the comment) and monetary damages. Because many agencies indemnify officials for actions taken in their official capacity, individual liability often becomes the agency’s bill, and legal fees alone are substantial even when the agency prevails.
Hatch Act violations carry termination and a five-year federal-service ban at the top of the range. Accessibility failures can produce Department of Justice complaints or private ADA lawsuits. Public records violations bring court-ordered compliance and monetary penalties that vary by jurisdiction.
The common thread is that a written, specific, consistently enforced social media policy is the best defense across all of these areas. Courts are more sympathetic to agencies that can show they trained employees on clear rules and followed them. Agencies operating without a policy, or with a vague one no one reads, absorb the full legal risk of every decision made by every employee with access to an official account.