In 2026, legal technology has burst out of the shadows with even movie stars trying to convince us they’ve been using these tools for years and some sports stadiums festooned with legal tech logos that 99% of the viewing audience won’t recognise (and frankly will never use).

With what we should assume are sizeable marketing budgets, look at the hyperscalers in legal tech. They have captured attention and headlines, with hardly a week going by without another organisation signing up for their services.

“Good for them,” should be the immediate response, but scratch beneath the surface, and a pattern emerges raising the question whether selection is being driven less by a rigorous and well-constructed evaluation of requirements, or more by market commentary, peer pressure, and FOMO.

It appears the rhetoric of “Everyone else is using it so….” or “no one gets fired for buying….” has slowly become a decision-making framework, and that’s not innovation. That’s imitation.

With the need to scale efficiently when incident volumes spike, AI has a real opportunity to change what breach notification review looks like. The catch is that it needs to be built into tested workflows with human QC at the center so that the speed gained from technology doesn’t come at the expense of defensibility.

Thoughtful GenAI workflows open the door for positive process improvement. The partnership between people and AI can give teams faster insight, and it can make repetitive work more consistent across a large data set.

One of the strongest conversations of the day came from Andy Cosgrove, Chief Strategy Officer at TCDI, and Deb Voronkov, VP of Data Governance at JPMorgan Chase & Co. The topic focused on information governance, which has always mattered, but matters a little more when organizations start feeding data into AI tools hoping for clean, useful answers on the other side.

During this interview-styled session, Deb shared what governance looks like in a large, complex data environment. She talked about the issues many teams are trying to manage right now:

• Growing data volumes
• Unclear ownership of AI systems and outputs
• Policies that don’t always reflect how people actually work

And in the middle of all these concerns, sensitive information still needs to be protected.

One of the bigger takeaways was that AI readiness begins long before a tool is selected. Organizations need to understand:

• What data they have
• Where that data lives
• Who owns, manages, and determines whether the rules around it still make sense.

Because at the end of the day, AI is only as useful and defensible as the information it’s given. If you can’t answer the questions above, you may find yourself dealing with significantly more problems than you anticipated once GenAI enters the picture. The tech may feel futuristic, but it still comes back to clean data.

When project variables change, the first step is to evaluate how those changes affect the current trajectory:

Some corporations and law firms are openly embracing AI because they see its potential to reduce costs and make legal work more efficient. These are the pilots. They’re evaluating use cases and rethinking workflows as they try to understand where this technology can help without creating a mess someone else has to clean up later.

Next, we have the stewards. These organizations are moving with far more caution, and that makes sense too. They see the opportunity, but also thinking about confidentiality, privilege, data security, defensibility, governance, and client expectations. Those concerns aren’t exactly minor footnotes. For those organizations, AI has to come with rules. It must be managed, and it has to fit within a risk framework that someone can defend.

And finally, we have the trailblazers. These organizations sit at the other end of the spectrum from the stewards and are actively requiring the use of AI in daily workflows. They’re setting expectations and looking at whether attorneys and legal professionals are leveraging this technology to its fullest potential.

And of course, there are a lot of organizations that sit somewhere between these groups, because trying to fit an entire industry into three neat categories simply isn’t realistic.

There is little doubt that in-house teams are paying attention. General counsel and legal ops leaders are actively engaging with AI, albeit in a very broad capacity. It’s all that can be discussed at conferences, and it’s beginning to be seen in briefings, in piloting tools, and in the exploration of vendor capabilities. However, this interest has not yet translated into marrying the known challenges with proven use.

Instead, most organisations remain in a discovery phase for AI usage and are asking the fundamental questions:

• Where does GenAI fit within existing legal process?
• What are the clear benefits to the business?
• What problems is it best suited to solve?
• How do we manage risk, accuracy, and accountability?

This reflects a pattern seen with previous technologies, such as bold promises and initial enthusiasm followed by a period of cautious validation before wholesale adoption.

A consistent thread I noticed throughout the forum was the shift toward greater ownership and operational maturity within in-house legal teams. In many sessions, I found parallels to the discussions we’ve been having during our LitForward Legal AI Roundtable sessions across the country.

For example, discussions emphasized standardizing processes, improving coordination with IT and cybersecurity teams, and rethinking relationships with outside counsel and service providers. The tone was pragmatic, with less theory and more real-world applications of how teams are actually adapting.

This was a refreshing change from events last year, highlight a shift from companies being wary of business partners’ use of AI to mandating it to maximize cost-effectiveness.