South Carolina: AI Ethics Guidance for Law Firms
Verified September 30, 2026
Summary
South Carolina has no bar ethics opinion on AI. The South Carolina Supreme Court's Interim Policy on the Use of Generative Artificial Intelligence (March 2025) governs Judicial Branch employees and reminds lawyers they are responsible for the accuracy of AI-assisted work. The Bankruptcy Court for the District of South Carolina has issued two practice notices on AI and Rule 9011 and sanctioned counsel for AI-fabricated citations in In re Richburg (2025). South Carolina added technology competence to the Rule 1.1 comments in 2019.
On this page
- What governs by default
- What federal courts in South Carolina require for AI use in filings
- What South Carolina state courts require for AI use in filings
- What AI-related rules are pending in South Carolina?
- How do malpractice carriers in South Carolina treat AI use?
- What does my South Carolina malpractice carrier ask about AI at renewal?
- What documentation should a South Carolina firm keep on file?
- Advisory and non-attorney rules (1)
- Decided AI hallucination cases (6)
- Applicable rules (reference)
Bottom line for a 5-50 attorney South Carolina firm: No South Carolina Bar opinion governs attorney AI use, so the operative framework is the Rules of Professional Conduct (Rule 407, SCACR), the 2019 technology-competence comment to Rule 1.1, and the Supreme Court’s March 2025 Interim Policy, which reminds lawyers “that they are responsible to ensure the accuracy of all work product.” In federal bankruptcy practice, the Bankruptcy Court has twice reminded the bar that Rule 9011 applies to AI-assisted filings, and it ordered AI-specific CLE for counsel who filed Copilot-generated citations. The most acute risk is hallucinated citations under Rule 3.3, with confidentiality under Rule 1.6 close behind. Firms have no safe-harbor opinion to point to.
Firms may wish to start with: a written AI use policy, a citation verification checklist for every court filing, and vendor due-diligence records for each AI tool in use.
What governs by default
As of 2026-09-30, the South Carolina Bar Ethics Advisory Committee has not issued an ethics advisory opinion on AI in legal practice. Its most recent opinions on the official index, EAO 26-01 and 26-02, address real estate closings and IOLTA bank selection. The operative framework is the South Carolina Rules of Professional Conduct (Rule 407, SCACR), with these rules most directly implicated by AI use:
- Rule 1.1 (Competence). By order of November 27, 2019, the Supreme Court amended the Rule 1.1 comment on maintaining competence (then Comment 6; renumbered Comment 8 by the Court’s November 17, 2021 order). The comment states that a lawyer “should keep abreast of changes in the law and its practice, including a reasonable understanding of the benefits and risks associated with technology the lawyer uses to provide services to clients or to store or transmit information related to the representation of a client.” The Court adopted “a modified version” of the Bar’s proposal, which drew on the ABA’s 2012 Ethics 20/20 amendments. Firms may wish to document how lawyers learn the known failure modes of AI tools (hallucinated citations, biased output, data retention) before using them. See the current rule text.
- Rule 1.6 (Confidentiality). The same 2019 order adopted Rule 1.6(c): “A lawyer shall make reasonable efforts to prevent the inadvertent or unauthorized disclosure of, or unauthorized access to, information relating to the representation of a client.” Before client information goes into an AI tool, firms may wish to evaluate whether the vendor trains on user inputs, its data retention policy, and whether its privacy architecture fits these obligations. Public free-tier tools that train on inputs present heightened risk.
- Rule 1.4 (Communication). South Carolina has not resolved whether AI use must be disclosed to clients. ABA Formal Opinion 512 is the closest available framework. It states that “client disclosure may be unnecessary” depending on the circumstances, but that lawyers must disclose GAI practices if a client asks, must obtain informed consent before inputting information relating to the representation into a self-learning tool, and must consult the client when GAI use bears on the fee or when its output “will influence a significant decision in the representation.”
- Rule 1.5 (Fees). Op 512 states that a lawyer “may not charge a client to learn about how to use a GAI tool or service that the lawyer will regularly use for clients,” and treats the cost of tools that function like general practice infrastructure as overhead.
- Rule 3.3 (Candor). Submitting AI-hallucinated citations or fabricated quotations to a tribunal can implicate Rule 3.3. The Supreme Court’s Interim Policy notes that generated content “may contain inaccuracies” and reminds lawyers and litigants “that they are responsible to ensure the accuracy of all work product and must use caution when relying on any output of Generative AI.”
- Rule 5.3 (Nonlawyer Assistance). Supervising attorneys remain responsible for reviewing AI-assisted work product produced by paralegals or staff.
What federal courts in South Carolina require for AI use in filings
As of 2026-09-30, no district-wide local rule, general order, or standing order on AI use in filings has been identified for the U.S. District Court for the District of South Carolina. The standing orders index contains no AI provision, and neither do the court’s Local Civil Rules (February 22, 2023) or Local Criminal Rules (revisions through July 17, 2019). Individual district judges post their own form orders on the court’s district judges page; those were not reviewed for this entry, so firms may wish to check the assigned judge’s form orders at the opening of a matter. District judges have addressed AI-style citation errors case by case, mostly in pro se matters: Bruce, Cherleatha B., Jeter, and Sevilla.
The U.S. Bankruptcy Court for the District of South Carolina has twice reminded the bar, in its “Select Court Observations on Practice” notices, that Fed. R. Bankr. P. 9011 applies to AI-assisted filings:
- July 2025. The notice stated that “it remains the filer’s ultimate responsibility to review and verify the accuracy of computer-generated content to ensure compliance.”
- April 2026. The notice stated that “attorneys are required to ensure the accuracy of all citations included in pleadings filed with the Court, regardless of whether those citations are generated by artificial intelligence.” It cites In re Richburg (Aug. 27, 2025), in which the court ordered counsel who filed Copilot-generated citations to complete three hours of ethics CLE, at least two of them on the ethical use of artificial intelligence.
Firms may wish to check the assigned judge’s posted forms at the opening of each federal matter. The operative defaults are Fed. R. Civ. P. 11 and Fed. R. Bankr. P. 9011.
What South Carolina state courts require for AI use in filings
Chief Justice John W. Kittredge issued the South Carolina Supreme Court Interim Policy on the Use of Generative Artificial Intelligence on March 25, 2025 (primary source PDF). As of 2026-09-30, no other Supreme Court order addressing AI was found. The policy applies to Judicial Officers and Employees of the Judicial Branch, which it defines to include “Justices, judges, attorneys, law clerks, administrative assistants, interns, externs, temporary employees, paralegals,” other employees and volunteers, and information technology professionals. It is not a rule governing private attorneys, but a separate section addresses lawyers and litigants by way of reminder.
Key provisions:
- Approved tools. Judicial Branch Officers and Employees may use only generative AI tools “approved by the Supreme Court or South Carolina Court Administration.” Tools the Branch procures or provides are deemed approved.
- Drafting. Employees may not use generative AI to draft memoranda, orders, opinions, or other documents “without direct human oversight and approval.”
- Reliance. AI content may not be used verbatim, assumed accurate, treated as the sole source of reference, or solely relied on in final decisions.
- Lawyers and litigants. Section (d)(1) states that the policy “does not specifically address the use of Generative AI by lawyers and litigants” and reminds lawyers and litigants that they are responsible for the accuracy of all work product. Section (d)(2), which applies to lawyers only, states that “Lawyers in particular must ensure that the use [of] Generative AI does not compromise client confidentiality or otherwise violate the South Carolina Rules of Professional Conduct, Rule 407, SCACR.”
The policy’s definition of generative AI names ChatGPT, Microsoft 365 Copilot, Grok, Gemini, Meta Chat, and Westlaw’s AI-Assisted Research and/or CoCounsel as examples; it does not designate them as approved tools. The policy remains in effect “until further Order of the Chief Justice or the Supreme Court.” In an August 15, 2025 news announcement the Court reminded “the Bench, Bar, and those who interact with our judicial system” of the policy and noted that it “may be revised in the future as needed.”
As of 2026-09-30, no statewide court rule requiring AI disclosure in filings was located.
What AI-related rules are pending in South Carolina?
The 2025-2026 General Assembly has considered several AI-related bills; none directly regulates attorney conduct.
- Enacted. S. 28 (Act No. 57 of 2025, AI-generated child sexual abuse material) and S. 29 (Act No. 58 of 2025, morphed images of minors) were signed on May 22, 2025.
- Chatbot Protection Act. S. 896 and H. 5138, companion bills, as introduced would have required chatbot providers to give “clear, conspicuous, and explicit notice to a user that the user is interacting with a chatbot rather than a natural person” and would have barred chatbot output implying that a chatbot is “a licensed legal professional.”
- On May 6, 2026, the Senate Labor, Commerce and Industry Committee reported S. 896 favorably with an amendment that replaced its text with a “Protecting Children from Chatbots Act.”
- The substitute is limited to operators with “fifty thousand or more monthly active users worldwide” and excludes chatbots used by a business entity “predominantly for customer service, order fulfillment, account management, or to provide users with information about commercial services or products provided by that entity.”
- No further action on either bill is recorded.
- Other tracking items, without direct attorney-conduct implications:
- S. 788 (AI in therapy or psychotherapy) passed the Senate on April 29, 2026 and was referred to a House committee.
- H. 4657 (Right to Compute Act) was prefiled December 16, 2025.
- S. 443 concerns review of health-coverage decisions from automated decision-making tools.
- S. 963 is a consumer-protection bill for interactions with AI systems.
- S. 225 is a Senate resolution supporting AI development.
The House committee with AI in its portfolio is the Regulations, Administrative Procedures, Artificial Intelligence and Cybersecurity Committee, chaired by Rep. Jeff Bradley.
How do malpractice carriers in South Carolina treat AI use?
No South Carolina-specific carrier AI guidance has been identified. The South Carolina Bar’s preferred insurance providers list names ALPS (“a South Carolina Bar preferred professional liability provider”), Pearl Insurance (The Hanover Insurance Group), and Swiss Re Corporate Solutions. ALPS’s August 21, 2025 guidance, which is national, states that LPL policies “typically do not exclude coverage for claims alleging negligence arising as a result of the use of generative AI,” while noting that “most LPL policies have exclusions related to intentional acts or breaches of confidentiality that are not the result of a negligent act.” Firms may wish to review current policy language for AI-related coverage or exclusions and confirm with the carrier or broker whether AI documentation affects renewal terms.
What does my South Carolina malpractice carrier ask about AI at renewal?
Whether South Carolina carriers’ application forms ask about AI was not checked for this entry. The substantive standard is the existing Rules of Professional Conduct (Rules 1.1, 1.6, 3.3, 5.3), the 2019 technology-competence comment, and ABA Formal Opinion 512 as persuasive authority. Firms may wish to be ready to produce a written AI use policy, vendor due-diligence records, citation verification logs for court filings, and a matter-level AI use log showing supervisory review, and to confirm application items directly with their broker.
What documentation should a South Carolina firm keep on file?
Month one (foundational)
- (Owner: managing partner + firm administrator) Written AI use policy. Approved tools list. Bar on inputting confidential client information into non-enterprise AI tools. Required human review of AI output before delivery. Supervisory chain for non-attorney AI use. Billing rules consistent with Rule 1.5 and Op 512.
- (Owner: litigation lead) Citation verification checklist. Required independent verification of every citation in any AI-assisted court filing or client document. This addresses Rule 3.3 candor risk, the Interim Policy’s reminder that lawyers are responsible for the accuracy of all work product, and the Bankruptcy Court’s April 2026 notice that attorneys must verify every citation regardless of whether AI generated it.
- (Owner: matter lead attorney) Judge-specific AI order check at the opening of each D.S.C. matter and, in bankruptcy matters, a check of the Bankruptcy Court’s practice notices. No district-wide rule exists, but individual chambers may impose their own requirements.
Months two and three (operational documentation)
- (Owner: firm administrator + outside IT) Vendor due diligence records for each AI tool. Capture: vendor name, product tier (free vs. enterprise), data retention and training policy, terms of service summary, last review date. This documents the Rule 1.6(c) reasonable-efforts inquiry.
- (Owner: firm administrator) CLE and training records tracking AI-related continuing education under the Rule 1.1 technology-competence comment. Rule 408, SCACR requires 14 CLE hours each reporting year, including 2 hours of legal ethics/professional responsibility; it has no AI- or technology-specific hour requirement.
- (Owner: managing partner) Supervision protocol covering AI-assisted work by associates, paralegals, and staff under Rule 5.3.
Months four to six (per-matter discipline)
These are recurring practices, not one-time projects. Each new matter exercises them.
- (Owner: matter lead attorney) Per-matter AI use notes. Capture the tool used, task performed, date, and the supervising attorney’s review of output, as a per-matter record of Rule 5.3 supervision and Rule 1.1 competence.
- (Owner: managing partner + billing partner) Engagement letter AI provisions disclosing AI use in general terms, with specific informed consent for matters where client information will go into a self-learning AI tool. Op 512 states that “merely adding general, boiler-plate provisions to engagement letters purporting to authorize the lawyer to use GAI is not sufficient.”
We are building practitioner resources for firms working through this list. The newsletter covers new resources as they ship.
Advisory guidance and non-attorney rules (1)
In force, but not an operative obligation on attorney filings: cautions and notices that restate existing duties, and rules that bind court personnel or parties rather than attorneys.
- South Carolina Supreme Court Interim Policy on the Use of Generative Artificial Intelligence , South Carolina (statewide judicial branch) ( Mar 2025 )
Decided AI hallucination cases in South Carolina (6)
Court, disciplinary and administrative matters addressing AI use in legal proceedings, including sanctions, warnings and responses that imposed no sanction; each case page states the outcome. Editorially flagged cases for South Carolina firms appear first with a "Why this matters" note; the remaining 3entries follow.
- In re Eric Chibueze Nwaubani
Why this matters: Unpublished Fourth Circuit disciplinary opinion on fabricated citations. Holds that existing Rule 8.4(d) is sufficient to sanction without an AI-specific rule.
- Bruce v. United States
Why this matters: Second of two D.S.C. AI-citation orders by Judge Lydon in a six-week window (Cherleatha B. Dec 29 2025; Bruce Feb 9 2026). Lydon chambers template.
- Sevilla v. Ross
Why this matters: Volume-based AI screening: Austin treats the count of pro se motions (80+) plus shotgun-pleading patterns as the AI flag, no specific fabricated cases needed.
- Cherleatha B. v. Bisignano , D.S.C. ( Dec 2025 )
- Jeter v. ADUSA Transportation LLC , D.S.C. ( Dec 2025 )
- In re Richburg , Bankr. D.S.C. ( Aug 2025 )
Applicable rules (reference)
How each rule applies to AI use in legal practice. Rule numbers reflect this state's own numbering where it diverges from the ABA Model Rules.
- Rule 1.1 : Competence
- In the firm AI policy, list approved tools and document what each one can and cannot do, where its data goes, and where it hallucinates. The competence duty rests on the lawyer, not the vendor.
- Rule 1.4 : Communication
- Decide once, by matter type, when AI use rises to a level the client should be told about. Document the threshold in the firm AI policy and reflect it in the engagement letter.
- Rule 1.5 : Fees
- Bill actual time, not pre-AI hourly time. If AI subscription costs are passed through to a client, disclose the basis in writing before the invoice issues.
- Rule 1.6 : Confidentiality
- Vet every AI tool that touches client data: data retention, training-data use, security posture, breach disclosure, and deletion on termination. No client-identifying input into a public-tier tool.
- Rule 3.3 : Candor Toward the Tribunal
- Personally read every cited case and verify every quoted authority before signing a filing. This is the rule behind every AI-hallucination sanction issued to date.
- Rule 5.3 : Nonlawyer Assistance
- Treat AI tools the way the firm treats paralegals: written supervision protocol, attorney review before client or court delivery, and named accountability per matter.