The Role of Firm Partners in Responsible AI Use

Your firm may not have formally adopted AI tools, but that doesn’t mean your associates aren’t already using them.

That raises an interesting question for firm partners: when AI starts entering the practice at the individual level, how should leadership respond?

What tools make sense for the firm?
What information should stay off-limits?
How do you preserve lawyer judgment while still making room for new technology?

Atty. Rogelio Torres, Jr. and Atty. Jerome Canlas, both partners at Cruz Marcelo & Tenefrancia (CMT), began thinking through these questions early. 

We spoke with them about how CMT approached the conversation, what they learned along the way, and how other firms can begin shaping their own approach to AI.

In this conversation, we discussed:

• Why AI adoption in the Philippines often starts with individual lawyers

• What firms risk by leaving AI use unaddressed

• How partners can set expectations and shape responsible use

• How to begin the conversation while AI regulation in the PH is still evolving

• Why individual lawyers remain responsible for how they use AI

The state of AI adoption in the PH: why is AI adoption being driven by end-users?

In larger U.S. firms, AI adoption is increasingly an organizational decision. A 2026 Bloomberg Law survey found that all 40 responding firms with 500 or more attorneys were already using legal-specific AI tools in 2025.

In the Philippines, adoption appears more bottom-up. I asked Attys. Torres and Canlas why younger lawyers seem to be adopting AI faster than the firms they work for.

Younger lawyers are adopting AI faster

Atty. Torres pointed partly to a generational gap. Experienced lawyers have established ways of working: “We have developed our own habits, our own approach.” Younger lawyers, meanwhile, “don’t have the same apprehension as us.” Atty. Canlas similarly described them as “digital natives” who adapt to new tools more readily.

That means experimentation can start at the individual level long before the firm has decided whether to formally adopt AI at all.

PH firms are navigating an evolving regulatory environment

The regulatory environment adds another layer.

In February 2026, the Supreme Court adopted the Governance Framework on the Use of Human-Centered Augmented Intelligence in the Judiciary, which sets broad principles around human oversight, accountability, transparency, confidentiality, and responsible AI use.

But firms still have relatively little specific guidance on how those principles should translate into day-to-day legal practice.

CMT chose not to wait for every rule to be settled before acting. Atty. Canlas pointed to other jurisdictions where lawyers had already been sanctioned after submitting AI-generated, hallucinated authorities to court. “We wouldn’t want to be the firm who submits hallucinated cases in our submissions,” he said. “We don’t want to be that.”

Atty. Torres described CMT’s approach as “more preventive rather than reactive.” The firm could establish expectations now, then adapt its policy as more specific Supreme Court guidance develops.

As Atty. Canlas put it:

“The march of AI is inevitable. So, we might as well be ready for it instead of suddenly being unprepared for it.”

From there, the question became less about whether AI should be used at all and more about what kind of use the firm was willing to stand behind.

The elephant in the room: what leadership risks by avoiding the AI conversation

CMT wrote its AI policy before formally choosing a platform. The partners knew AI use could already be happening inside the firm, and that ignoring this reality created two risks:

• Immediate ethical and compliance issues

• A deeper risk to the quality and identity of the firm itself

The risk of ethical and compliance issues

Confidentiality is one of CMT’s earliest concerns. “We wanted to avoid shadow AI use,” Atty. Canlas said. “We were afraid that the associates might upload confidential information on the cloud.”

But sensitive information doesn’t only appear in uploaded documents. A prompt itself can reveal legal strategy. “If I type, ‘Give me the five latest cases on this specific doctrine,’” he explained, “if it comes into the wrong hands this person would have an idea what case strategy will be.”

Without firm-wide guidance, lawyers are left to make those calls individually: which tools to use, what information to enter, and what level of risk is acceptable.

The risk of losing your firm’s identity

For Atty. Torres, the deeper question is what unguided AI use does to the practice itself. “If we don’t address it, we run the risk of losing our identity.”

As a teaching firm, CMT sees developing younger lawyers as part of its responsibility. Ignoring AI could make the firm’s approach obsolete; relying on it indiscriminately could undermine the judgment and habits it wants associates to develop. 

As Atty. Torres put it:

“If we don’t address it, if we don’t harness it in the right direction, then we run the risk of making ourselves obsolete and making our brand non-existent.”

The challenge is to make AI consistent with the kind of practice the firm wants to preserve.

For leadership, that means thinking beyond productivity and considering whether AI use still reflects the kind of lawyers, and the kind of firm, they’re trying to build.

Atty. Torres tied that responsibility directly to the firm’s reputation:

“If they take pride in their name being on the wall, they have to take pride in how services under their name are rendered.”

That also means avoiding an all-or-nothing approach. Firms should not refuse to use AI at all, but they should not use it for everything either. As Atty. Torres explained, lawyers are still required to make judgment calls, and AI is not meant to substitute for them.

Culture flows from the top: how leadership shapes a firm’s culture on AI usage

For Atty. Canlas, AI governance should be partner-driven because it reflects how the firm wants to practice, what standards it wants to uphold, and where AI should fit into that work.

1. Start with what you want to achieve and avoid

“The partner should know what you want to achieve. The partner should know what you want to avoid.”

That became the starting point for CMT’s own policy. The partners wanted AI to improve efficiency and quality without creating “the perils of cognitive decline” or “the perils of sensitive information leaking.”

For partners beginning the same conversation, that means first agreeing on a few basic questions:

• What do we want AI to help our lawyers do better?

• What parts of our practice should AI never replace?

• What risks are we unwilling to accept?

• What standards should remain true regardless of the technology being used?

As Atty. Canlas shared: “We drafted the AI policy so that we know how to best utilize AI without sacrificing our values as a firm.”

2. Ask associates what’s already happening

Partner-driven doesn’t mean partner-only. 

Associates are often the ones experimenting with AI first, so they may have the clearest view of how it’s already entering day-to-day work.

Atty. Torres said firms should use that experience:

“The data can also be taken from the associates. They can be asked what their experiences are, what their concerns are.”

For leadership, that can mean asking directly:

• Which tools are lawyers already trying?

• What are they using them for?

• Where are they uncertain about what is allowed?

• What risks or limitations have they encountered?

3. Model the behavior you want to see

For Atty. Torres, responsible AI use also has to be modeled by leadership. Critical thinking isn’t something associates develop simply because they’re told to do it: “They have to practice it. They have to see us doing it.”

That same thinking shaped CMT’s approach to AI. Rather than simply discouraging associates from experimenting with new tools, the partners saw an opportunity to learn alongside them and show what responsible use could look like in practice.

“If they’re going to use it, let’s use it all together. Let’s learn all together.”

That approach is consistent with how Atty. Torres described the firm’s identity:

“We’re still professional, we’re still competent, but we can still be more if we use this tool properly.”

He connected that identity to the firm’s day-to-day mentoring:

“Every year we bring in new lawyers. I always work with the first-year lawyers. I always kid with them that I’m their batchmate. I never get old. It’s really something that we do enjoy, and that’s our identity and our strength.”

The responsibility of the user: AI doesn’t make the judgment call for you

Firm-wide policies can set expectations, but they can’t make every decision for the individual lawyer.

Atty. Torres emphasized that confidentiality remains the lawyer’s responsibility. If there’s doubt about whether information should be uploaded, his advice is conservative: understand the platform’s data policies first, and if the doubt remains, don’t take the risk.

“At the end of the day, the burden is on the lawyer,” he said. “Our obligation to be competent doesn’t outweigh our obligation to hold client confidences.”

That responsibility starts with the query itself. “Don’t be lazy. Don’t just upload without thinking,” Atty. Torres said. “There’s a way to pose the query and still maintain confidentiality.”

The same applies to the output. AI can retrieve information, suggest an argument, or produce a draft, but the lawyer still has to verify whether it’s correct, applicable, and appropriate for the client.

“It’s always the responsibility of the lawyer. It’s never the AI.”

AI adoption is already a leadership question - "don't choose to ignore"

Cruz Marcelo & Tenefrancia chose to engage with that reality early, making room for AI while being clear about what the firm wasn’t willing to compromise.

For Atty. Torres, that’s the role of leadership: you don’t need to have every answer, but you do need to help set the direction.

“Choose what to use, but don’t choose to ignore.”