At SlatorCon Remote March 2026, Slator’s Head of Consulting, Alex Edwards, spoke with experts from security system provider Ajax Systems and the privacy-first web browser Brave, in a partner panel supported by Crowdin that focused on localization for privacy-first companies.
During the panel, Olena Azanova, the Localization Team Lead at Ajax Systems, and Andy Andersen, International Growth Product Manager at Brave, shared their perspectives on localization in privacy-first sectors, the current role of AI in their workflows, and the future of AI translation in their field.
Edwards began with a question on how localization differs in a privacy-focused environment. Perhaps counterintuitively, Azanova noted the difference for Ajax is that many more internal stakeholders, such as R&D, product, marketing, and legal teams, all participate in the pipeline before content reaches her team.
She noted that, “it doesn’t mean that each product, or each update, or each localization step has to be approved by each team. But the amount of stakeholders on one device, or one new feature, is just enormous.”
Andersen took a high-level perspective, highlighting Brave’s need to address varying regulations and privacy concerns from end-users in different cultures. He encapsulated this by saying, “privacy is a bit of a sliding scale depending on where you are and what you’re localizing for.”
This led the panel into a discussion of cultural and legal differences between markets. Azanova stressed the importance of pre-research before entering a market, highlighting not only differing regulations in different countries, but even the word-level differences that come up when localizing.
Using the term “compliance” as an example, she noted that “in English, compliance is widely known. It’s understandable. But in different countries, especially in Arabic-speaking countries, and in more Eastern European countries, we have to describe this word because it’s understandable, of course, but it’s not that widely used as it is in English.”
Andersen highlighted research challenges facing privacy-first companies, saying, “when you’re a privacy-first company, you also don’t really have a whole lot of information about your users. […] you have to conceptualize a lot of these things and figure out what’s the best use case for your end-users rather than getting it directly from a lot of data like a regular B2C company would.”
AI for Privacy-First Localization
Noting that both Ajax and Brave use Crowdin as their Language Technology Platform (LTP) partner, Edwards asked the speakers about their experience using the platform.
Azanova praised the company’s security focus, saying, ”I’m a huge advocate for Crowdin. We’ve been using it for years […] we can’t use just everything that we want to use because of the security issues. Crowdin is really great there.”
For his part, Andersen highlighted Crowdin’s flexibility and responsiveness, calling their support “incredibly helpful in a time like this when there are a lot of changes in the space.”
As the conversation shifted to AI translation, both panelists agreed the technology is valuable but “not quite there yet.” Azanova mentioned this is particularly the case in privacy-first sectors because of the sensitivity of their content.
At the same time, she described Ajax’s content-tiering strategy of “gradually deciding” what content requires human-first translation, and what can be AI translated, then post-edited. However, she stressed that her team’s workflows remain human-led, saying, “not a single piece of content is going to the website or to the user without human review.”
Andersen honed in on persisting challenges like brand voice, saying, “AI might help you to move faster, but is it really giving you the personality and the brand touch? […] I think I’m not personally seeing it.” He further noted that while AI translation broadly performs well in high-resource languages, it falls short when targeting specific locales like French for Canada or Arabic for Morocco.
Has AI Hit a Plateau?
Looking towards the future, Edwards asked the panelists whether the industry had reached an AI development plateau, and what that would mean for their product strategy.
Azanova summed it up, stating, “short answer is yes.” She then elaborated, “I think for our industry, in security, we will reach a plateau. I don’t think there will ever be this AI and localization and translation world where we can just do everything automated without a human in line at all.”
She emphasized the need for human accountability, saying, “I know that I can be 100% sure that everything that goes to our website, or to the user, or to the app, is more or less correct. […] I know that we have someone accountable for that, which AI cannot be at this point.”
Andersen agreed regarding the AI plateau while acknowledging that there will continue to be “certain tasks” that AI tools can handle. He expressed more concern around integrating AI into workflows, noting, “I think in general the biggest question mark that we’re going to figure out in the industry is more logistical. It’s kind of ‘how do the tools integrate better with these technologies, and how do you implement that into your workflow at scale?’”