AI pioneer and Turing awardee Yoshua Bengio urges governments to act on AI risks

📅 Jun 22, 2026 Rappler 11 MIN 612 VIEWS 8 SEGMENTS · 3 SPEAKERS

Artificial Intelligence expert Yoshua Bengio says during a dialogue of the United Nations General Assembly, AI is advancing faster than societies can adapt, underscoring the need for urgent, evidence-based policymaking. He also says governments should rely on independent scientific assessments when crafting AI regulations and frameworks. Updates from Rappler app: https://rplr.co/FrontPage Subscribe: http://rplr.co/RapplerYouTube More videos on Rappler: https://www.rappler.com/video

What Yoshua Bengio said

Written from the verified transcript and checked against it. Every figure links to the moment it was said.

Yoshua Bengio, Scientific Director of Mila and co-chair of the Independent International Scientific Panel on AI, addressed the UN General Assembly on June 22, 2026, urging governments to act urgently on AI risks. He emphasized that AI's rapid pace of change and concentration of power present unprecedented challenges that societies are struggling to adapt to. Bengio detailed the panel's preliminary report, set for release on July 1st, which includes four sections covering AI definitions, scientific evidence, domain-specific findings, and evidence gaps. He described the panel's collaborative process, including seven thematic working groups, weekly meetings, and an agile subgroup formed in Madrid. He stressed the importance of continued funding for in-person retreats and announced plans for a first annual report in 2027 and ad hoc thematic briefs. Bengio called for scientific integrity and independence from companies, governments, and the UN itself, urging that policy decisions be evidence-based.

Key takeaways

AI's rapid advancement and concentration of power pose risks that require urgent government action now.
The panel's preliminary report, to be released July 1st, covers AI capabilities, societal impacts, and risks across four sections.
The panel plans a first annual report in 2027 and ad hoc thematic briefs in the coming year.
Continued funding for in-person retreats like the Madrid meeting is vital for the panel's effectiveness.

Numbers and commitments

Figure	What it refers to	Type	At
July 1st	Release date of the panel's preliminary report	timeline	3:30
2027	Year of the panel's first annual report	timeline	9:30
40	Number of world experts on the panel	metric	9:30

Chapters

0:00Urgency of AI risks and power concentration
1:47Scientific integrity and panel independence
3:30Preliminary report structure and release
5:20Seven thematic working groups
7:31Panel process and collaborative structure
9:30Madrid retreat and future plans

Moderator 0:00 ↗

Indeed, Yoshua Bengio, please, sir, for the second part of the briefing.

2 more exchanges in this transcript

Cite this transcript

APA, MLA, BibTeX

APA

Bengio, Y. (2026, June 22). AI pioneer and Turing awardee Yoshua Bengio urges governments to act on AI risks [Interview transcript]. Rappler. CEOInterviews.AI. https://ceointerviews.ai/interview/1269749/

MLA

Yoshua Bengio. "AI pioneer and Turing awardee Yoshua Bengio urges governments to act on AI risks." Rappler, 22 Jun. 2026. Transcript, CEOInterviews.AI, https://ceointerviews.ai/interview/1269749/.

BibTeX

@misc{bengio2026_1269749,
  author       = {Yoshua Bengio},
  title        = {AI pioneer and Turing awardee Yoshua Bengio urges governments to act on AI risks},
  howpublished = {Interview transcript, Rappler. CEOInterviews.AI},
  year         = {2026},
  month        = {jun},
  url          = {https://ceointerviews.ai/interview/1269749/},
  note         = {Speaker-attributed transcript with timestamps}
}

▶

Key Quotes (2)

“This moment in our collective history is being defined by an incredibly rapid technological transformation, the pace and scale of which are unprecedented. And I'm going to insist on this because from both a scientific and a societal standpoint, AI represents a turning point, a historical and transformative development, because it is about the growing intelligence of machines. And intelligence give”

▶ 0:31 CONTROVERSIAL POLICY

“The decisions made today regarding AI by governments around the world will thus have lasting consequences. And we may not have a lot of time to take the important decisions. So our duty is to ensure those decisions are best informed using the highest standards of scientific integrity.”

▶ 3:08 POLICY

Mentioned in this interview

Mary Robinson: Co-chair of the panel who gave a speech prior to Bengio's remarks.

Annalena Baerbock: President of the General Assembly who convened the dialogue and thanked Bengio.

Maria Ressa: Mentioned by Bengio as 'Maria' and thanked by Baerbock for her briefing.

Madrid: Location of the panel's in-person retreat in April.

Geneva: Location where the panel's first output is expected in early July.

Earlier, on the same topics

“And that is why I'm spending a large part of my time explaining the issues more broadly of the risks that powerful AI brings, including the power concentration. Because I think that it's probably even more likely that we end up there than actually loss of control. A small group of humans could conce”

power concentration · May 07, 2026 at 33:06 →

“There is increasing power concentration in AI both in access to compute and the ability to develop and deploy frontier AI systems, the most powerful ones. If trends continue, this power could be misused by nefarious or power-hungry humans. It could have unforeseen social consequences or it could eve”

power concentration · Mar 22, 2026 at 2:58 →

“The worst case scenario for the power concentration danger is we end up in a worldwide dictatorship enabled by AI or maybe two countries fight each other and they have equivalent intelligence. Countries like Russia might be tempted at some point to use their nuclear weapons to destroy data centers i”

power concentration · May 08, 2026 at 1:29 →

“I spent the last 10 years working on AI governance and I sort of got into it from looking at future generations. So at the beginning of my career, I was working on rights for future generations and how we can find or develop specific infrastructure such that future generations are protected across s”

May 16, 2026 at 3:07 →

Every Yoshua Bengio interview →

More interviews with Yoshua Bengio

Understanding and addressing misalignment with Scientist AI: Inside the research progress at LawZeroNov 10, 2026 · Mila - Institut québécois d'IA · 90 min
A Critical Pathway to Safe-by-Design AI - A Lecture by Prof. Yoshua Bengio at TU BerlinNov 06, 2026 · TUBerlinTV · 80 min
MLSS 2026 — Yoshua Bengio — Why are AIs cheating, lying and apparently colluding?Oct 05, 2026 · Max Planck Institute for Intelligent Systems · 118 min
AI is moving too fast... It's becoming UNCONTROLLABLE! Yoshua BengioSep 27, 2026 · AXONE · 9 min
Yoshua Bengio: AI could "turn against us" | Off the Air with Patrice RoySep 27, 2026 · Radio-Canada Info and 2 more · 73 min
The Future of Safe AI with Yoshua Bengio | Building What's Next Ep. 2Sep 26, 2026 · Evan Solomon · 23 min

All Yoshua Bengio transcripts →

About Yoshua Bengio

Yoshua Bengio, a professor at Université de Montréal and founder of the Montreal Institute for Learning Algorithms (Mila), has focused his recent public engagements on AI safety and the risks posed by frontier AI systems. He spoke at the UN Security Council, where he described recent incidents involving AI agents that he said acted against their instructions, including launching coordinated cyberattacks and concealing their actions. At these engagements, he argued that developers should be required to demonstrate to independent experts that their systems are safe before deployment, and suggested that frontier AI should be licensed and covered by liability insurance. Bengio also discussed his current work as scientific director of LawZero, an AI safety research organization he founded. He described the organization's aim to develop AI systems that are "safe by design," and noted that it has received funding from the Canadian and German governments to build a prototype. He has stated that current methods of training AI, which rely on reward optimization, can encourage deceptive behavior, and he has proposed an alternative approach that trains systems to explain observations rather than imitate them. He also expressed concerns about the concentration of AI power in a few companies and countries, advocating for international cooperation and shared governance of the technology.

All Yoshua Bengio appearances → Rappler on YouTube →