TL;DR AI use in legal filings is not banned, but courts are making one thing clear: lawyers remain responsible for accuracy. Bay Area law firms should treat AI-assisted drafting, research, and citation work as supervised legal work that requires verification, documentation, and clear internal controls.
Artificial intelligence is already part of legal work. Attorneys may use it to summarize records, review discovery, organize timelines, draft arguments, or test language before a filing is finalized. Used carefully, AI can help law firms work more efficiently. Used casually, it can create serious professional and operational risk.
The Northern District of California has already given Bay Area law firms a practical warning. In a February 2026 standing order, Magistrate Judge Susan van Keulen stated that generative AI tools are not prohibited, but attorneys and self-represented parties must personally confirm the accuracy of AI-generated content. The order also says that signing a submission containing AI-generated content certifies that the content has been personally verified.
That matters because the issue is no longer whether AI is allowed. The issue is whether law firms have a process to make sure AI-assisted work is accurate, secure, and properly supervised.
What Does the Northern District of California Say About AI Use in Legal Filings?
The Northern District of California standing order does not ban AI. Instead, it places responsibility where it belongs: on the person submitting work to the court.
The order states that counsel is responsible for complete and accurate representations in any submission to the court, including filings, demonstratives, evidence, and oral argument. It specifically names generative AI tools such as ChatGPT, Claude, and Gemini as tools that may be used, but only if the attorney or party personally confirms the accuracy of the content.
The most important point is simple. If a lawyer signs a filing that contains AI-generated content, including AI-generated citations, that signature certifies that the content was personally verified.
The order also says the court will impute AI errors to the attorney or party whose signature appears on the document. In plain English, a lawyer cannot blame the tool. The responsibility stays with the filer.
Why Should Bay Area Law Firms Care?
Bay Area law firms operate in a region where AI adoption is moving quickly. Clients may expect faster turnaround. Attorneys may face pressure to produce more in less time. Staff may experiment with public AI tools because they seem helpful and easy to use.
That creates a blind spot.
The risk is not just that AI can be wrong. The risk is that AI can be wrong while sounding polished, confident, and legally plausible. For a law firm, that can mean a fake citation, an incorrect summary, a misstated legal standard, or a document that looks complete but has not been properly verified.
A single inaccurate citation can damage credibility with the court. A confidential client document pasted into a public AI tool can create privacy and privilege concerns. A lack of internal policy can make it unclear who reviewed the work before it was filed.
This is why AI use in legal filings should not be treated as a personal preference. It should be treated as a firmwide operational risk.
Did You Know?
California Rule of Court 10.430 requires courts that allow generative AI use by court staff or judicial officers to adopt a generative AI use policy by December 15, 2025. Those policies must address confidential information, verification, hallucinated output, bias, disclosure, and compliance with applicable laws and ethical rules. [Source: Judicial Branch of California]
This is aimed at courts, not private law firms. But law firms should pay attention to the direction California courts are taking.
California Standard 10.80 also warns judicial officers not to enter confidential, personal identifying, or nonpublic information into public generative AI systems. It further says judicial officers should take reasonable steps to verify AI-generated material and correct erroneous or hallucinated output.
That language maps directly to law firm risk. If courts are creating AI rules around confidentiality and verification, law firms should not wait until a mistake forces the issue.
What Is the Difference Between AI Assistance and AI Reliance?
AI assistance means using AI as a tool while humans remain in control. A lawyer may use AI to organize facts, outline a draft, summarize a long document, or brainstorm possible issues. The output is reviewed, checked, corrected, and approved by a qualified person.
AI reliance is different. It happens when a firm treats AI output as accurate because it looks professional. That is where the danger grows.
Legal work requires more than fluent writing. It requires judgment, context, accurate sources, confidentiality, and accountability. AI can help with some parts of the process, but it does not replace the lawyer’s duty to verify.
The better question is not, “Can we use AI?” The better question is, “What must happen before AI-assisted work is trusted?”
The Real Problem Is Shadow AI
Shadow AI happens when employees use AI tools without the firm’s approval or oversight. In a law firm, this could look like a paralegal pasting deposition notes into a chatbot, an associate using AI to summarize cases, or an assistant using AI to rewrite client-facing language.
The intent may be harmless. The risk is not.
If the firm does not know which tools are being used, it cannot know what data is being entered, whether prompts are being stored, whether outputs are being verified, or whether client confidentiality is being protected.
For law firms, shadow AI creates two separate problems. The first is accuracy risk, especially around citations, quotes, facts, and legal conclusions. The second is confidentiality risk, especially when client data is entered into public or consumer-grade tools.
Read more in “Shadow AI Is Already in Your Business. Can You See It?”
What Should Law Firms Do Before AI Reaches a Filing?
A law firm does not need a complicated AI program to reduce risk. It needs clear rules, consistent review, and basic visibility.
First, identify which AI tools are already being used inside the firm. This includes free public tools, browser extensions, legal research platforms, Microsoft 365 AI features, and vendor tools.
Second, define what information may never be entered into public AI systems. That should include privileged information, confidential client data, personal identifying information, sealed materials, protected health information, financial data, employment records, and sensitive litigation materials.
Third, require human verification before any AI-assisted material is used in a court filing. Every case citation, statute, quotation, factual statement, and legal proposition should be checked against a trusted legal source. AI should not be treated as the authority.
Fourth, assign responsibility. If staff use AI to help prepare a draft, the supervising attorney should know how the output was created and how it was verified. If the work is filed with a court, the signer owns the accuracy.
Finally, train the team. A policy sitting in a folder is not enough. Attorneys and staff need plain-English training on hallucinations, confidentiality, prompt handling, and when AI use should be escalated.
How Does This Connect to Managed IT and Cybersecurity?
AI governance is now part of law firm technology risk. It connects directly to access controls, data protection, Microsoft 365 permissions, endpoint security, browser controls, vendor review, and employee training.
A law firm may have strong cybersecurity tools and still have AI risk if staff are copying sensitive information into unapproved platforms. A firm may also have excellent attorneys and still face filing risk if no one has defined how AI-assisted legal research must be verified.
This is where managed IT and cybersecurity support can help. The goal is not to make legal decisions for attorneys. The goal is to create safer systems around the tools the firm already uses.
That may include reviewing Microsoft 365 sharing permissions, identifying risky browser extensions, documenting approved AI tools, strengthening identity security, and helping leadership build a practical AI use policy.
Read more in “How Law Firms Can Reduce IT Risk Without Slowing Down Their Practice.”
Explore “Modern IT for Law Firms: Building a Smarter, Safer Practice.”
Learn how Managed IT Services can support secure day-to-day operations.
Practical AI Policy Checklist for Law Firms
A practical law firm AI policy should answer several basic questions.
- Which AI tools are approved for firm use?
- What types of client or case information are prohibited from being entered into AI tools?
- Who is responsible for verifying AI-generated citations, quotes, facts, and legal conclusions?
- When must AI use be disclosed to a supervising attorney, client, court, or opposing party?
- How are prompts, outputs, drafts, and verification steps documented when needed?
- Who reviews new AI-enabled vendor tools before the firm adopts them?
- How are nonlawyer staff trained and supervised when using AI?
These are not theoretical questions anymore. Courts are already addressing AI-generated content, and California’s legal ethics conversation is moving in the same direction.
FAQ
Is AI use in legal filings banned in the Northern District of California?
No. Magistrate Judge Susan van Keulen’s standing order does not prohibit generative AI use. It says attorneys and self-represented parties must personally confirm the accuracy of AI-generated content used in court submissions.
Who is responsible if AI creates a fake citation?
The attorney or party who signs the filing remains responsible. The Northern District standing order says AI errors will be imputed to the attorney or party whose signature appears on the document.
Does this apply to all California law firms?
The specific standing order applies to matters before that judge. However, the broader lesson applies to every law firm using AI: verify the work, protect confidential information, and do not treat AI output as authoritative without review.
What is an AI hallucination?
An AI hallucination is false or unsupported information generated by an AI system. In legal work, this may include fake cases, inaccurate citations, fabricated quotes, or misstated legal rules.
Should law firms stop using AI?
Not necessarily. Law firms should use AI carefully, with written policies, approved tools, verification steps, confidentiality controls, and attorney supervision.
About Professional Computer Concepts
Professional Computer Concepts (PCC) is a trusted Managed IT and Cybersecurity provider serving the Bay Area for over 20 years. We help small and midsize businesses simplify their IT, strengthen security, and modernize operations. Explore our services:
Managed IT Services | Cybersecurity | Cloud Solutions
From PCC’s Desk
AI can help law firms move faster, but speed is not the same as accuracy. For legal work, the final responsibility still belongs to the people signing, reviewing, and managing the work.
If your law firm is using AI or suspects employees may be using it without clear rules, now is the time to create a safer process. Let’s talk.