---
title: Who Answers for AI? The Governance Gap | ABBYY
description: Watch ABBYY’s Andrew Pery and Roman Kilun unpack new research from 1,200 IT leaders on how organizations approach accountability, data governance, risk management, and AI ROI.
image: https://digital.abbyy.com/hubfs/templates/sharing%20-%20new/other/logo.png
---

[![ABBYY logo](https://digital.abbyy.com/hubfs/abbyy-logo.svg)](https://www.abbyy.com)

- [Overview](https://digital.abbyy.com/who-answers-for-ai-governance-gap-webinar-on-demand#intro)
- [Speakers](https://digital.abbyy.com/who-answers-for-ai-governance-gap-webinar-on-demand#speakers)
- [Watch the webinar](https://digital.abbyy.com/who-answers-for-ai-governance-gap-webinar-on-demand#form)

On-demand webinar | 41 minutes

# Who Answers for AI When It Gets It Wrong?

New ABBYY research from 1,200 IT leaders shows AI adoption is outpacing governance. Hear what it means, and how to turn policy into controls you can prove.

[Watch the webinar](https://digital.abbyy.com/who-answers-for-ai-governance-gap-webinar-on-demand#form)

## Closing the AI Accountability Gap: From Pilots to Production

AI is moving from pilots to production faster than most organizations can govern it. The result is a gap between what AI can do and who is accountable when it doesn’t do what you expect.

In this conversation, Andrew Pery and Roman Kilun walk through the findings of ABBYY’s State of Intelligent Automation Report, a survey of 1,200 IT leaders and line-of-business professionals across 20 industries and six markets. They explain where organizations stand on AI governance, where accountability breaks down, and what it takes to scale AI with trust, control, and compliance.

![ds-1322-thumbnails-for-assets-on-abbyycom4-1](https://digital.abbyy.com/hubfs/Imported%20sitepage%20images/ds-1322-thumbnails-for-assets-on-abbyycom4-1.jpg)

### What you’ll learn

Who owns AI outcomes.

Responsibility is split across providers, deployers, and users, and ownership is often unclear. Learn how to define it, document it, and make it operational. 

How the EU AI Act divides responsibility.

Understand the different obligations of providers and deployers, and why modifying a model can change yours.

How to move from policy to evidence.

See the three pillars of operational governance (risk, controls, and assurance) and why auditors, customers, and regulators will ask you to prove them. 

What effective human oversight looks like.

Learn why “a human somewhere in the process” is not enough, and what to log, escalate, and review. 

Why data and documents decide your AI results.

Hear how data quality shapes both governance and ROI, and how DocLang aims to carry governance rules inside the document itself. 

Why governance speeds AI up.

Hear how leading with governance helps teams make decisions faster and move from experimentation to wider deployment. 

## Meet the speakers

![Andrew-Pery](https://digital.abbyy.com/hubfs/templates/speakers/Andrew-Pery.png)

###### Andrew Pery

AI Ethics Evangelist, ABBYY

View bio

![Close](https://digital.abbyy.com/hubfs/close.svg)

![Andrew-Pery](https://digital.abbyy.com/hubfs/templates/speakers/Andrew-Pery.png)

###### Andrew Pery

AI Ethics Evangelist, ABBYY

 Andrew Pery is an AI Ethics Evangelist at global intelligent automation company ABBYY. Andrew has more than 25 years of experience spearheading tech management programs for leading global technology companies. He specializes in intelligent document process automation, process intelligence, AI governance, data privacy, and ethical AI practices. He frequently shares his expertise through media commentary, original publications, and features on platforms like the AI Pulse Podcast, where he unpacks the complexities of AI legislation and societal impacts.Andrew holds a Master of Laws degree with Distinction from Northwestern University Pritzker School of Law, is a Certified Data Privacy Professional and a certified AI Auditor. He is a member of the American Bar Association, the International Association of Privacy Professionals, and dedicates his time to advocating for responsible technology adoption.

![Roman Kilun](https://digital.abbyy.com/hubfs/Speakers/Roman%20Kilun.png)

###### Roman Kilun

Chief Compliance Officer, ABBYY

View bio

![Close](https://digital.abbyy.com/hubfs/close.svg)

![Roman Kilun](https://digital.abbyy.com/hubfs/Speakers/Roman%20Kilun.png)

###### Roman Kilun

Chief Compliance Officer, ABBYY

 Roman Kilun is Chief Compliance Officer (CCO) at ABBYY, a leader in AI-powered document processing for the intelligent enterprise.Roman served as Corporate Counsel at ABBYY for nearly 15 years and was appointed to CCO in 2024. As Corporate Counsel, Roman led the modernization of ABBYY’s legal function, unifying and standardizing global systems, contracting processes, and dispute resolution, and managed the legal side of the company’s transition to a SaaS subscription model. As CCO, he serves as ABBYY's voice on the company's AI Risk Management Policy and cloud assurance certifications, and is responsible for the strategic direction, implementation, and monitoring of the company’s compliance program across global operations.He holds an undergraduate degree from the University of California, Berkeley and a law degree from UC Law San Francisco. A former professional cyclist, he still rides and enjoys tinkering on vintage scooters and cars.

#### Governance is the bridge between AI adoption and accountable outcomes.

 Find out who answers for AI in your organization before something goes wrong.  Watch the webinar

### FAQ

What is the AI governance gap?

  It’s the distance between how fast organizations adopt AI and how well they can govern it. In ABBYY’s survey, 54% said adoption has outpaced their ability to govern it. 

Who is responsible when an AI system produces a wrong or harmful outcome?

It depends on the system and the use case, but organizations should be able to answer, and document, who owns decisions, risks, and controls across providers, deployers, and users. Shared responsibility is the best policy. 

Does AI governance impact adoption?

  Providers place AI systems on the market and carry obligations around risk management, documentation, and data governance. Deployers use those systems and must follow the provider’s instructions. Substantially modifying a model can give a deployer obligations similar to a provider’s.

What does effective human oversight look like?

  It means knowing when human review is required, why a decision was escalated, who reviewed it, and what action was taken, with an auditable record.

Why does data quality matter for AI governance?

AI can only produce reliable outcomes when its inputs are reliable. In ABBYY’s research, 20% name data quality as the biggest barrier to AI ROI.

ABBYY’s hybrid approach pairs purpose-built AI with LLMs to turn unstructured documents into accurate, validated data, so your AI starts from inputs you can trust.

###### SUBSCRIBE FOR INSIGHTS

[Sign up](https://www.abbyy.com/subscribe/)

Connect with us

<https://www.facebook.com/ABBYYsoft/>

<https://www.linkedin.com/company/abbyy>

<https://twitter.com/ABBYY_Software/>

<https://www.youtube.com/user/ABBYYVIDEOS/videos>

<https://discord.com/invite/NefYHFKcXt>

<https://www.tiktok.com/@abbyy.global>

© 2026 ABBYY. All rights reserved.

- [Contact us](https://www.abbyy.com/company/contact-us/)
- [Legal](https://www.abbyy.com/legal/)
- [Third-party trademarks](https://www.abbyy.com/legal/third-party/)
- [Privacy policy](https://www.abbyy.com/privacy/)
- [Cookie policy](https://www.abbyy.com/cookie-policy/)
- [Cookie management](https://digital.abbyy.com/who-answers-for-ai-governance-gap-webinar-on-demand#cookiebot)
- [Vulnerability disclosure](https://www.abbyy.com/vulnerability-disclosure-guidelines/)
- [Visit abbyy.com](https://www.abbyy.com)