From “AI Anxiety” to Action: Lessons applying AI to Training from GES 2026

According to benchmarking polls by Ethena in late 2025, most compliance leaders are using AI for personal productivity, but fewer than 20% are running limited pilots that go beyond individual use. The most frequent concern on the mind of the CCO? “I feel like I’m already behind.”

That AI anxiety is quite real. But as Roxanne Petraeus, CEO of Ethena, reminded the crowd at the 2026 Global Ethics Summit: “Everyone is behind, which means no one is behind.”

With the help of Dianne Ramos, Head of Ethics & Customer Complaints at Guardian Life, Roxanne led From Prompt to Practice: A Hands-On Workshop for the AI-Curious CCO, a “hands-on keyboards” workshop to show how compliance teams can move from curiosity to actually applying Artificial Intelligence to compliance.

For those who couldn’t make this workshop at the Summit, here are four of its key takeaways.

1. “Gen Z-ing” the Message: From Text to Short Form Video

The core challenge for modern compliance isn’t just delivering information, it’s capturing attention. Training has to pass the vibe check with incoming workforces (especially younger employees) if it hopes to engage them.

At Guardian, Dianne’s team is moving from text-heavy slides in Guardian’s Ethical Management Training and transforming them into short form videos.

There are three primary ways to accomplish this transition to video:

• DIY Recording: A 60-second “selfie-style” video from a leader often feels more human and authentic than a polished corporate production.
• Professional Actors: Ethena’s actors can record these short-form videos themselves rather than in-studio, which increases speed and reduces cost.
• AI Video Generation: The breakthrough for speed and scale is generative AI video. Compliance teams can now turn a written policy update into a video in minutes rather than weeks.

During the workshop, Ethena demonstrated GenAI tool HeyGen to build a video in real-time based on a script generated from Ethena’s training. While the technology is powerful, a few tips to keep in mind for those starting out:

• AI Video Fatigue: Keep your videos under 60 seconds and mix them with other formats. Try to focus on one singular, actionable takeaway per video.
• Mix up the format: A static AI avatar can feel uncanny, so use different camera angles (e.g., zooming in for emphasis) and switch avatars between videos to keep the viewer’s eye engaged.
• Audio post-production: AI voices can sometimes sound robotic. Tweaking the audio in a post-production tool can help.
• Specificity is Success: Treat the AI like a “green intern.” Don’t just prompt: “Make a video about gifts.” Be specific: “Create a 60-second video for a global sales team explaining that local gifts under $50 are fine, but cash is always prohibited. Keep the tone friendly but firm.”

2. The New “Build vs. Buy” Paradigm: Speed, Control, and Consistency

A major focus of the session was how the “build versus buy” decision for compliance training is being fundamentally disrupted. Dianne shared how Guardian Life’s transition to a decentralized L&D model (with embedded teams across various business lines) necessitated a new approach to content creation:

• Faster Development Cycles: To keep pace with business needs, the time from concept to deployment has to shrink. AI allows teams to move away from blank-page syndrome, generating initial drafts and structures in seconds.
• Greater Control and “Redlining”: Traditionally, off-the-shelf compliance training is difficult to customize. Guardian needed to be able to “redline” content quickly by modifying existing training to fit internal needs without having to build from scratch.
• Tone and Voice Consistency: Despite decentralization, Guardian’s training had to maintain a consistent organizational tone. This is where AI helps ensure that the voice remains professional and cohesive.

The decision to build or buy is no longer as binary as it once was. AI can now adapt existing training with realistic, company-specific information (including tailored scenarios) which allows compliance teams to produce highly customized, specialized content that might otherwise be too resource-heavy to create in-house. By leveraging these tools, CCOs can finally produce both high-speed and high-relevance training that provides a higher return on investment for the organization’s most valuable resource: its employees’ time.

This customization doesn’t need to stop at the company level. AI makes it easier to build training per department or geography. By combining an off-the-shelf training, smart prompting, and a policy document, you can quickly “remix” training down to individual profiles:

• Risk-Based Tailoring: If your risk assessment shows a spike in bribery risks in a specific region, AI can take your standard anti-corruption module and rewrite the scenarios to reflect the local culture and norms.
• Role-Specific Context: A software engineer doesn’t need the same Conflicts of Interest training as a procurement lead. AI can help to pivot the tone and examples by moving from financial interests scenarios for procurement to moonlighting scenarios for devs, without requiring a human to rewrite every word.
• Workforce Nuance: Beyond just roles, AI helps adjust the complexity and reading level of content to better match the workforce, whether they are frontline hourly workers or executive leadership.

3. The Death of “Click-Next”: AI-Powered Interactivity

To keep employees from tuning out, AI is moving us beyond the tired formula of talking heads videos, multiple-choice quizzes, and text and images. AI enables the bespoke creation of many emerging interactive learning and training formats, including:

• Text Message Exchanges: Instead of reading a static story, employees see a simulated SMS thread. The “texts” between colleagues lay out a scenario and they can then choose how they might respond. AI is great at generating these “text message scenarios” instantly from a simple prompt and/or policy document.
• Matching Puzzles & Sorting: To break the pattern and ensure learners encounter something new, interactive drag-and-drop puzzles provide a way for employees to categorize items based on whether they’re acceptable or not (e.g., Acceptable Gift vs. Bribe). This gamified approach turns a passive experience into an active one.
• AI Grading (Free-Text Responses): Rather than being faced with a multiple choice question where many employees may randomly pick answers until they are correct, with AI Grading, employees are asked to explain why a situation is problematic in their own words or how they might handle a tricky situation. Ethena’s AI then grades the response in real-time based on specific criteria, returning a response to the employee in seconds. This forces employees to think through gray areas, a critical outcome for manager-level training.

4. The Global Voice: Localization using AI Translations

For global organizations, language translations can post a significant hurdle to ensure a compliance program is relevant and can be easily consumed by a multinational workforce, making translated content available and accessible is critical.

All too often, compliance teams struggle to translate their content. They may complete a given course and then translate into the 20 languages they need, but then a simple change to the source content has the ripple effect of needing to re-translate all content.

To avoid the manual translation nightmare, AI tools can now assist with fast and inexpensive translations that have become remarkably accurate:

• Seamless Dubbing: Any video with a speaker delivering a message in English can instantly be translated into Portuguese, using their actual voice.
• Human in the loop: Easy editing tools ensure humans can review translations and correct the minor mistakes to get the translation “just right.”
• Learner Preference: Some compliance programs lock learners into the most common language of their locale. But for Spanish speakers in the U..S or an expat in Brazil, learners should be able to choose their preferred language.

Ethena then showed how it has embedded ElevenLabs, a leading provider of AI audio and translation services, directly within its own tools, which automatically translates all training (including AI-generated training). Text can be played as audio in any language, and learners can choose their preferred language.

Conclusion: From Anxiety to Agency

The throughline of this session wasn’t any single feature or tool, it was a shift in mindset. The compliance teams seeing real results with AI today aren’t the ones who wait for perfect AI or try to boil the ocean. They’re the ones who picked a real problem, ran a pilot, and started learning.

Dianne at Guardian Life made the case plainly: four weeks of content development compressed down to one. Saving time is great, but the real question is what this now enables you to do. When AI handles the first draft and policy translation into multiple languages, compliance professionals can do the work only humans can do: building relationships, navigating gray areas, and ensuring the training actually reflects the culture and values of the organization.

This marks a transition from creator to editor for the compliance professional. It’s not that AI is putting jobs at risk, but rather that it enables us to do all the things we could only have dreamed of before. We can now build training for the software engineer and the procurement lead. We can respond to a regional bribery risk spike in days, not months. And, we can finally close the gap between what employees need to know and what we have the bandwidth to build.

AI isn’t going to run your compliance program. But it will give your team leverage they’ve never had before. All it takes is putting your hands on the keyboard.