Shared Responsibility
Industry and the Future of AI Safety
Industry and the Future of AI Safety
Industry and the Future of AI Safety
Industry and the Future of AI Safety
Panel & Evening Reception
PAST EVENT
AI Safety Connect and DGA Group convened senior industry leaders, policymakers and researchers to examine how the private sector can help shape frontier AI safety practices, norms and standards. The event took place hours after India's Minister of Electronics and IT unveiled the New Delhi Frontier AI Commitments at the main Summit.
Welcome & Opening Remarks
Mr. Nicolas Miailhe
Co-Founder, AI Safety Connect
Mr. Nicolas Miailhe
Co-Founder, AI Safety Connect
Mr. Cyrus Hodes
Co-Founder, AI Safety Connect
Mr. Cyrus Hodes
Co-Founder, AI Safety Connect
Mr. Paul Triolo
Partner, DGA Group
Mr. Paul Triolo
Partner, DGA Group
Ms. Sunita Patnaik
Partner, DGA Group
Ms. Sunita Patnaik
Partner, DGA Group
Cyrus Hodes welcomed the New Delhi Frontier AI Commitments but pressed for deeper engagement:
Cyrus Hodes welcomed the New Delhi Frontier AI Commitments but pressed for deeper engagement:
Cyrus Hodes welcomed the New Delhi Frontier AI Commitments but pressed for deeper engagement:
"Commitments around data insights and multilingual evaluations, while important, do not yet address the hard questions: how safety decisions are actually made at the frontier, how risks are assessed before deployment and how we hold each other accountable when things go wrong. We need more."
"Commitments around data insights and multilingual evaluations, while important, do not yet address the hard questions: how safety decisions are actually made at the frontier, how risks are assessed before deployment and how we hold each other accountable when things go wrong. We need more."
"Commitments around data insights and multilingual evaluations, while important, do not yet address the hard questions: how safety decisions are actually made at the frontier, how risks are assessed before deployment and how we hold each other accountable when things go wrong. We need more."
Honorable Minister Shri Duddilla Sridhar Babu
Honorable Minister Shri Duddilla Sridhar Babu
Minister for Information Technology, Electronics and Communications, Industries and Commerce and Legislative Affairs, State Government of Telangana
Minister for Information Technology, Electronics and Communications, Industries and Commerce and Legislative Affairs, State Government of Telangana
Minister Sridhar Babu positioned the Global South as an active author of safety standards:
Minister Sridhar Babu positioned the Global South as an active author of safety standards:
Minister Sridhar Babu positioned the Global South as an active author of safety standards:
"The Global South is not a place that is waiting for safety standards to be written elsewhere. We are writing them here, in real time, for 40 million people."
"The Global South is not a place that is waiting for safety standards to be written elsewhere. We are writing them here, in real time, for 40 million people."
"The Global South is not a place that is waiting for safety standards to be written elsewhere. We are writing them here, in real time, for 40 million people."
He announced Telangana's intention to serve as a regulatory co-builder alongside initiatives like AI Safety Connect.
He announced Telangana's intention to serve as a regulatory co-builder alongside initiatives like AI Safety Connect.
He announced Telangana's intention to serve as a regulatory co-builder alongside initiatives like AI Safety Connect.
Shri Sanjay Kumar
Shri Sanjay Kumar
Special Chief Secretary, Industry, Commerce, IT and Communications, State Government of Telangana
Special Chief Secretary, Industry, Commerce, IT and Communications, State Government of Telangana
Kumar framed the challenge directly:
Kumar framed the challenge directly:
Kumar framed the challenge directly:
"AI is something where everybody has to come together. Every country has to come to the same platform to discuss and debate because technology and AI doesn't understand any boundary."
"AI is something where everybody has to come together. Every country has to come to the same platform to discuss and debate because technology and AI doesn't understand any boundary."
"AI is something where everybody has to come together. Every country has to come to the same platform to discuss and debate because technology and AI doesn't understand any boundary."
He outlined Telangana's data exchange platform, which anonymises public data for startups and researchers while preserving privacy.
He outlined Telangana's data exchange platform, which anonymises public data for startups and researchers while preserving privacy.
He outlined Telangana's data exchange platform, which anonymises public data for startups and researchers while preserving privacy.
Deciding at the Frontier — How AI Gets Deployed, Adopted, and Governed
Mr. Combiz Abdolrahimi
Global Head of Government Affairs & Public Policy at ServiceNow
Mr. Ravi Aurora
Senior Vice President, International Government Affairs
Ms. Joslyn Barnhart
Senior Research Scientist, Google DeepMind
Ms. Terah Lyons
Managing Director & Global Head of AI & Data Policy at JPMorgan Chase
Moderator
Mr. Amlan Mohanty
Mr. Amlan Mohanty
Fellow, Carnegie India
Fellow, Carnegie India
The panel explored how advanced AI systems are increasingly being deployed and tested in live environments at scale. Consequential safety judgments, such as when systems are ready to launch, how risks are assessed and how reporting and disclosure are handled, are often shaped through internal decision-making alongside evolving regulatory expectations. Senior leaders from ServiceNow, Mastercard, Google DeepMind and JPMorgan Chase examined where current industry approaches are proving robust and where uncertainties persist, including how companies navigate the tension between competitive pressure to ship and the growing expectation of responsible deployment.
The panel explored how advanced AI systems are increasingly being deployed and tested in live environments at scale. Consequential safety judgments, such as when systems are ready to launch, how risks are assessed and how reporting and disclosure are handled, are often shaped through internal decision-making alongside evolving regulatory expectations. Senior leaders from ServiceNow, Mastercard, Google DeepMind and JPMorgan Chase examined where current industry approaches are proving robust and where uncertainties persist, including how companies navigate the tension between competitive pressure to ship and the growing expectation of responsible deployment.
The panel explored how advanced AI systems are increasingly being deployed and tested in live environments at scale. Consequential safety judgments, such as when systems are ready to launch, how risks are assessed and how reporting and disclosure are handled, are often shaped through internal decision-making alongside evolving regulatory expectations. Senior leaders from ServiceNow, Mastercard, Google DeepMind and JPMorgan Chase examined where current industry approaches are proving robust and where uncertainties persist, including how companies navigate the tension between competitive pressure to ship and the growing expectation of responsible deployment.
Governing Frontier AI for Secure Deployment — Interoperable Standards in a Fragmented World
Mr. Austin Mayron
Acting Director of the U.S. Center for AI Standards and Innovation
Mr. Chris Meserole
Executive Director at the Frontier Model Forum
Mr. Chris Meserole
Executive Director at the Frontier Model Forum
Mr. Michael Sellitto
Head of Global Affairs, Anthropic
Ms, Amanda Craig
General Manager, Responsible AI Public Policy, Microsoft
Moderator
Mr. Paul Triolo
Mr. Paul Triolo
Partner, DGA Group
Partner, DGA Group
Discussant
Mr. Christian Troncoso
Mr. Christian Troncoso
Principal, Global AI Policy, Amazon Web Services
Principal, Global AI Policy, Amazon Web Services
The panel examined where cross-border divergences in AI safety approaches are creating uncertainty and what practical forms of coordination could move beyond dialogue toward operational compatibility. Anthropic's Michael Sellitto argued that regulatory frameworks must scale with capability:
The panel examined where cross-border divergences in AI safety approaches are creating uncertainty and what practical forms of coordination could move beyond dialogue toward operational compatibility. Anthropic's Michael Sellitto argued that regulatory frameworks must scale with capability:
The panel examined where cross-border divergences in AI safety approaches are creating uncertainty and what practical forms of coordination could move beyond dialogue toward operational compatibility. Anthropic's Michael Sellitto argued that regulatory frameworks must scale with capability:
"When you're riding a bicycle, you don't need an airbag. When you're driving a car that can go 100 miles an hour, you do."
"When you're riding a bicycle, you don't need an airbag. When you're driving a car that can go 100 miles an hour, you do."
"When you're riding a bicycle, you don't need an airbag. When you're driving a car that can go 100 miles an hour, you do."
Chris Meserole drew on other industries to make the case that interoperable standards are achievable:
Chris Meserole drew on other industries to make the case that interoperable standards are achievable:
Chris Meserole drew on other industries to make the case that interoperable standards are achievable:
"We've done this in many other industries, whether it's aviation or others. In the AI space we're still very early in that process, but that's something we're trying to work towards long term."
"We've done this in many other industries, whether it's aviation or others. In the AI space we're still very early in that process, but that's something we're trying to work towards long term."
"We've done this in many other industries, whether it's aviation or others. In the AI space we're still very early in that process, but that's something we're trying to work towards long term."
Closing Remarks
The co-founders closed with three asks of industry participants: work toward shared language on safety within and across organisations; bring governments, including middle powers, into standard-setting as co-authors rather than consumers and return to the next conversation with something concrete — a commitment, a pilot or a metric.
Nicolas Miailhe identified the evening's central finding:
Closing Remarks
The co-founders closed with three asks of industry participants: work toward shared language on safety within and across organisations; bring governments, including middle powers, into standard-setting as co-authors rather than consumers and return to the next conversation with something concrete — a commitment, a pilot or a metric.
Nicolas Miailhe identified the evening's central finding:
Closing Remarks
The co-founders closed with three asks of industry participants: work toward shared language on safety within and across organisations; bring governments, including middle powers, into standard-setting as co-authors rather than consumers and return to the next conversation with something concrete — a commitment, a pilot or a metric.
Nicolas Miailhe identified the evening's central finding:
Closing Remarks
The co-founders closed with three asks of industry participants: work toward shared language on safety within and across organisations; bring governments, including middle powers, into standard-setting as co-authors rather than consumers and return to the next conversation with something concrete — a commitment, a pilot or a metric.
Nicolas Miailhe identified the evening's central finding:
"There is a gap between what we aspire to and what we can accomplish to implement standards of safety and trust for frontier AI. The good news is that the people working to close that gap, in coordination with multiple sectors and across borders, are in the room with us tonight."
"There is a gap between what we aspire to and what we can accomplish to implement standards of safety and trust for frontier AI. The good news is that the people working to close that gap, in coordination with multiple sectors and across borders, are in the room with us tonight."
"There is a gap between what we aspire to and what we can accomplish to implement standards of safety and trust for frontier AI. The good news is that the people working to close that gap, in coordination with multiple sectors and across borders, are in the room with us tonight."
Photo Gallery
Photo Gallery Attendees can access the gallery with the link and password provided in your follow-up email.
