How Law Firms Can Build AI Search Authority

Here is a test you can run in under two minutes. Open ChatGPT, Claude, or Perplexity. Type: "Recommend a corporate law firm in Jakarta for cross-border M&A." Then try the same for Singapore.

Count the names. Count the detail. Count the confidence level of the response.

For Singapore, the AI will name specific firms, cite specific partners, and reference specific practice areas. For Jakarta, you will get vague category descriptions and maybe a directory link. Not because Jakarta lacks excellent law firms. It does not. But because those firms have not built the infrastructure that allows AI to verify and cite them.

This is not a marketing problem. It is an entity infrastructure problem. And it is fixable.

What AI agents need before they name a law firm

AI systems do not recommend firms the way a colleague does. A colleague trusts reputation and personal experience. AI trusts what it can verify from multiple independent sources.

For a law firm, verification requires a specific chain. The firm itself must be a verifiable entity. The individual attorneys must be verifiable entities. The practice areas must be explicitly declared in machine-readable format. And there must be published evidence connecting those attorneys to those practice areas through actual legal analysis.

Break any link in that chain, and the AI either skips you or hedges its recommendation with "you may want to verify this independently." Which is a polite way of saying it does not trust you enough to stake its own credibility on naming you.

As I wrote in Schema Markup Is Not Technical. It's Strategic, structured data is not a developer task. It is the primary language AI agents use to understand who you are. For law firms, this matters more than for most industries, because legal services are high-trust, high-consequence decisions where AI systems apply extra scrutiny before recommending.

The three pillars of law firm AI authority

After auditing dozens of law firm websites across Southeast Asia, I have identified three specific areas that determine whether a firm gets cited by AI. Not traffic. Not rankings. Citation.

1. Partner entity verification

Every named partner at a firm needs to exist as a verified entity in AI's understanding of the world. This means more than a bio page on the firm website. It means the partner's identity is consistent and verifiable across multiple independent platforms.

What that looks like in practice: a LinkedIn profile that matches the firm bio exactly. Bar association listings that confirm admission and standing. Published articles or legal analysis under the partner's name on the firm domain and external publications. An ORCID profile if the partner has published academic or research work. A Google Scholar profile if they have written anything citable.

Most Indonesian law firms have partner bios that read well as prose but carry zero structured data. The bio says "Budi Santoso specializes in corporate restructuring with 20 years of experience." That is nice for a human reader. For an AI agent trying to verify whether Budi Santoso is a real attorney who actually practices corporate restructuring, it is nearly useless. There is nothing to cross-reference. Nothing to confirm.

Singapore firms do this differently. Not because they are thinking about AI (most are not). But because the professional ecosystem demands it. Partners at Rajah & Tann or Allen & Gledhill have profiles on the firm site, on LinkedIn, on legal directories like Chambers and Legal 500, on bar association databases, and often on university affiliation pages. That density of consistent information across independent sources is exactly what AI needs to treat a person as a verified entity.

2. Practice area schema

A law firm's practice areas need to be declared in LegalService schema on the website. Not just written in prose. Declared in JSON-LD that AI crawlers can parse without interpretation.

LegalService is a schema.org subtype of LocalBusiness specifically designed for legal providers. It lets you declare practice areas, jurisdictions, office locations, and the attorneys connected to each service area. When implemented correctly, it creates an unambiguous machine-readable map of what the firm does, where it does it, and who does it.

Here is the gap. I audited 20 top-tier law firms in Jakarta. Two had any form of LegalService schema. Two. And both implementations were incomplete, missing practice area specificity and attorney connections.

In Singapore, the number was closer to 15 out of 20. Not because Singapore firms are obsessed with schema markup. But because their web development ecosystem considers structured data a standard deliverable, not an optional add-on.

3. Published legal analysis

AI systems need something to cite. A practice area page that says "we handle corporate restructuring" gives the AI nothing to reference. A published analysis titled "Key Considerations in Indonesian Cross-Border M&A After the 2025 Investment Law Amendments" gives the AI a specific, citable, authoritative piece of content tied to a verified attorney at a verified firm.

This is where the gap between Singapore and Indonesia becomes most visible. Singapore firms publish legal alerts, client briefings, and analysis pieces on their own domains. They distribute through legal publications like IFLR, Asian Legal Business, and Chambers. Their attorneys write for law journals and contribute to bar association publications.

Indonesian firms tend to keep their analysis internal. Client memos stay in email. Briefings stay in PDF decks. The institutional knowledge is deep, but it is invisible to any system that relies on publicly accessible, attributed content.

I have seen this pattern before. In Mengapa Perusahaan Singapura Muncul di AI Saat Kamu Tidak, I documented the same gap across industries. Law firms are not special in this regard. They are just a particularly acute example, because the stakes of AI citation are unusually high in legal services.

The entity chain: how it connects

These three pillars are not independent. They form a verification chain that AI systems follow when deciding whether to cite a firm.

The chain works like this. An AI agent receives a query about cross-border M&A firms in Jakarta. It looks for firms with LegalService schema declaring M&A as a practice area and Jakarta as a served jurisdiction. It checks whether the firm's identity is consistent across external sources. It looks for published analysis from verified attorneys at that firm on M&A topics. If all three links hold, the firm gets cited. If any link breaks, the firm gets skipped.

This is not a ranking algorithm. It is a trust verification process. And the process is binary. Either the chain holds, or it does not.

Singapore vs. Indonesia: the data

I audited the top 20 law firms by headcount in Singapore and Jakarta across six entity infrastructure metrics. The results are not subtle.

The numbers tell a clear story. Singapore firms have built entity infrastructure as a byproduct of their professional ecosystem. Indonesian firms have not, because their ecosystem never required it. Both ecosystems produced excellent lawyers. Only one produced lawyers that AI can verify.

The specific gaps:

LegalService schema: 75% adoption in Singapore vs. 10% in Indonesia. This is the most fixable gap. A competent developer can implement LegalService schema in a day.

Attorney schema: 80% vs. 5%. Singapore firms mark up individual attorney profiles with Person schema connecting credentials, bar admissions, and practice areas. Indonesian firms use unstructured HTML bios.

Published analysis: 90% vs. 25%. This is the hardest gap to close quickly, because it requires sustained content production. But even here, many Indonesian firms have existing internal analysis that could be adapted for public publication.

ORCID profiles: 45% vs. 5%. Partners at Singapore firms with academic or research output register ORCID profiles. This is rare in Indonesia outside of academia.

Consistent NAP: 85% vs. 30%. Name, address, phone consistency across platforms. Basic, but Indonesian firms frequently have mismatched information between their website, Google Business Profile, and legal directories.

FAQ structured data: 60% vs. 8%. Structured FAQ sections that AI systems can extract and cite directly in response to legal queries.

Entity infrastructure checklist for law firms

If you are a managing partner reading this and thinking "where do we start," here is the prioritized checklist. The protocol in How to Audit a Competitor's Entity Infrastructure applies here. Start by auditing what you have, then close the gaps in order.

Total effort for the critical layer: roughly two weeks of focused work. Not a year-long digital transformation project. Two weeks. The gap between being invisible to AI and being citable by AI is measured in weeks, not years.

Why law firms are uniquely vulnerable

Legal services have characteristics that make this problem especially urgent.

High trust threshold. AI systems apply stricter verification standards to legal recommendations than to, say, restaurant recommendations. The consequences of citing an unqualified or non-existent attorney are severe. So AI agents demand more verification before naming a law firm. If you have not built the verification infrastructure, you are filtered out more aggressively than firms in lower-stakes industries.

Competitor density. Every major jurisdiction has dozens of qualified firms for any given practice area. AI does not need to take risks on firms it cannot verify when there are five verified alternatives available. The decision is not "should we recommend this firm?" It is "we have six verified firms and this unverifiable one. Skip the unverifiable one."

Client behavior shift. In-house counsel at enterprises are using AI tools during the firm selection process. Not to make the final decision, but to build the initial long list. If your firm is not on the AI-generated long list, you are relying entirely on existing relationships and referrals. That works until it does not.

Cross-border complexity. International clients doing due diligence on Indonesian law firms are asking AI systems for verification. If the AI cannot confirm your practice areas, partner credentials, and jurisdictional standing from independent sources, the international client moves to a firm the AI can verify. Often a Singapore firm with an Indonesia desk.

The schema implementation that matters most

If I had to pick one single action for an Indonesian law firm to take this month, it would be implementing LegalService schema on the firm website. Here is what a minimal correct implementation looks like:

{
  "@context": "https://schema.org",
  "@type": "LegalService",
  "name": "Firm Name",
  "description": "Full-service Indonesian law firm...",
  "url": "https://firmname.co.id",
  "telephone": "+62-21-XXX-XXXX",
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "Jl. Sudirman...",
    "addressLocality": "Jakarta",
    "addressCountry": "ID"
  },
  "areaServed": [
    {"@type": "Country", "name": "Indonesia"},
    {"@type": "City", "name": "Jakarta"}
  ],
  "hasOfferCatalog": {
    "@type": "OfferCatalog",
    "name": "Legal Services",
    "itemListElement": [
      {"@type": "Offer", "itemOffered": {
        "@type": "Service", "name": "Corporate M&A"
      }},
      {"@type": "Offer", "itemOffered": {
        "@type": "Service", "name": "Banking & Finance"
      }}
    ]
  },
  "sameAs": [
    "https://www.linkedin.com/company/firmname",
    "https://chambers.com/law-firm/firmname"
  ],
  "founder": {
    "@type": "Person",
    "name": "Managing Partner Name"
  }
}

This is not complex. It is not expensive. It is a structured declaration of what the firm already is. The only reason most Indonesian firms do not have it is that nobody told them they needed it.

What happens when you do nothing

The trajectory is predictable. AI search adoption among enterprise buyers is accelerating. In-house legal teams are already using AI tools for preliminary research and vendor discovery. Within 18 months, AI-assisted firm selection will be standard practice for cross-border transactions involving Indonesia.

Firms that have built entity infrastructure will be on the long list. Firms that have not will be invisible to every procurement process that starts with an AI query. And increasingly, that is most of them.

The firms that act now have a compounding advantage. Entity authority builds over time as AI systems encounter and re-verify the same consistent information. Starting six months from now means starting behind firms that started today.

A note on compliance

Some managing partners will worry about regulatory compliance when publishing legal analysis. This is a valid concern but not a blocking one. Bar association rules (including Indonesia's PERADI and HKHPM codes of conduct) generally permit educational content and legal commentary. They restrict specific claims about case outcomes and client guarantees.

The solution is straightforward. Publish analysis that explains legal frameworks, summarizes regulatory changes, and provides general guidance. Do not publish specific case outcomes or guarantee results. This is the same standard that Singapore firms follow, and it is sufficient for building AI citation authority.