Require AI labs to publish safety evaluations before deploying frontier models

ai-risk transparency ai-safety ai-regulation regulations research-policy policy ai-governance ai-ethics

Share this with your representatives, journalists and friends

Key insights

Independent evaluation appears on both sides of the debate, but the disagreement is whether audits should be mandatory and public or voluntary and more limited.
Many supportive quotes describe publication as summaries, model cards, transparency reports, or technical reports, suggesting that transparency is often framed short of full raw-data disclosure.
Scope is a recurring design issue: several supporters want obligations aimed only at the largest frontier developers and most advanced models, not the broader AI market.
Supporters often connect pre-deployment publication to a wider lifecycle regime that also includes post-deployment monitoring, framework updates, and incident reporting.
The quotes come from governments, researchers, civil society groups, standards bodies, trade associations, and some frontier AI companies, showing that support and skepticism arise from multiple constituencies.

Arguments for

Pre-release proof of safety

Proponents say frontier models can create serious chaos, cyberattacks, loss of control, and other catastrophic harms, so developers should have to show systems are safe before release. Several quotes add that launches should be withheld, blocked, or reversed when adequate mitigations are not in place or high safety standards are not met.

This is a four-alarm fire. We need to take action now. These frontier models could have the power to create serious chaos and harm without adequate regulations. This bill, while not perfect, is an important, bipartisan step in the right direction. We need a bill like this that ca…

Suhas Subramanyam · Jul 23, 2026 · full quote
Frontier model AI systems have the potential to be used for many good things. However, due to their massive computing power, they also could cause catastrophic events, such as cyberattacks or the system evading control by developers or users. In the absence of federal oversight o…

Kwame Raoul · Jul 6, 2026 · full quote
Make companies prove their systems are safe before release — at every step, the burden of proof should be on the companies, not the consumer or the public.

Delia Ramirez · Jun 4, 2026 · full quote
Frontier AI models, like airplanes, should be required to go through technical testing and auditing, and their release should be blocked or reversed as a threat to public safety if they do not meet high standards of safety.

Dario Amodei · Jun 2026 · full quote
There's much more to do, including giving CAISI the authority to create the standards we need to know whether a frontier model is safe before it ships, not after something goes wrong.

Lori Trahan · Jun 25, 2026 · full quote
If... we were to find that we were reaching a critical capability level and we didn't have the appropriate mitigations, then we would not be launching.

Helen King · Jul 16, 2026 · full quote

Independent verification builds trust

Supporters argue that safety should be verified by independent testers rather than accepted on the word of developers who profit from deployment. Mandatory independent testing, certification, and publication of results are presented as ways to build trust and accountability for the public, developers, customers, and regulators.

We need regular mandatory independent safety testing and oversight, mandatory disclosure of security incidents, and international cooperation to keep people safe from absolute disaster.

Greg Casar · Jul 23, 2026 · full quote
Both must rest on independent third-party oversight, trust verified by third parties, not asserted by the developers who profit from them.

Sam Gregory · Jul 7, 2026 · full quote
SB 315 is the strongest AI safety law in the country. It builds on the frameworks California and New York have already passed and goes one step further by requiring independent audits, so the public doesn’t have to take AI companies at their word. Illinois has set the new standar…

Sunny Gandhi · Jul 6, 2026 · full quote
Innovation alone is not enough. Independent testing and certification are essential to building trust and confidence across the AI life cycle.

Jason Chan · Jul 17, 2026 · full quote
Our Advanced AI Framework proposes mandatory independent safety testing for frontier models, transparency requirements, and government authority to block or recall dangerous AI deployments.

Anthropic · Jun 12, 2026 · full quote
The Institute will establish safety standards, build open-source evaluations that AI developers can run against their models, independently test AI products, and publish the results to provide transparency and accountability.

Common Sense Media · May 5, 2026 · full quote

Shared standards and coordination

Many proponents frame publication requirements as part of common baselines, uniform rules, and internationally coordinated evaluation methods. They argue that clear standards and interoperable reporting can avoid fragmented rules and support global trust in systems with cross-border reach.

The window to get AI policy right is closing, and Congress can't afford to miss it. I supported Representatives Trahan and Obernolte's initial draft so we could develop a better bill that reflects how this technology works and where it's headed. Thanks to input from researchers, …

Scott Peters · Jul 23, 2026 · full quote
We need much stronger co-ordination of testing of frontier models before they go into wide circulation. This needs to be co-ordinated internationally because no country can seal itself off from the cross-border nature of systems that are prevalent today. Such co-ordinated testing…

Andrew Bailey · Jul 14, 2026 · full quote
AI safety is a challenge that no single nation can meet on its own. Through this memorandum of understanding, the Republic of Korea and Canada will jointly advance evaluation methodologies and measurement science for frontier AI, laying the groundwork for internationally interope…

Myuhng-Joo Kim · Jul 8, 2026 · full quote
We need common baselines for frontier systems, common methods to evaluate and verify risks, and common resolve that systems with global reach must meet standards worthy of global trust.

António Guterres · Jul 6, 2026 · full quote
Strengthen disclosure from private companies who are developing and deploying AI globally. Build a coherent international assurance ecosystem.

Rebecca Finlay · Jul 7, 2026 · full quote
The General-Purpose AI Code of Practice’s Safety and Security Chapter require providers to conduct model evaluations also through adequately qualified independent external evaluators. These reports will then support AI Office’s regulatory activities since providers are required t…

European Commission · 2026 · full quote

Voluntary disclosure leaves gaps

Some supporters say deployment is outpacing safety evaluation, labs publish too little empirical evidence, and no one is forcing third-party review. They argue mandatory publication is needed because models can interfere with evaluations and some systems are rushed to market without basic testing or with testing results ignored.

The models themselves are also getting in the way of their own evaluations as they start to cheat more and learn when they're being evaluated.

Lawrence Chan · Jul 24, 2026 · full quote
The latest AI Agent Index reveals the widening gap between the pace of deployment and the pace of safety evaluation. Most developers share little information about safety, evaluations, and societal impacts.

Leon Staufer · Feb 19, 2026 · full quote
Developers publish broad, top-level safety and ethics frameworks that sound reassuring, but are publishing limited empirical evidence needed to actually understand the risks.

Leon Staufer · Feb 19, 2026 · full quote
Equally concerning are reports that new LLM models are being sped to market without basic safety testing, or with the results of that testing ignored.

Office of the New York State Attorney General · Jan 15, 2026 · full quote
There’s no one forcing them to work with third-party experts to make sure that things are safe and secure.

Miles Brundage · Jan 15, 2026 · full quote
I have spent the past year working with technical colleagues to document what it would take for an AI company to meaningfully reduce the risk and harms of these systems. It turns out, none of them are particularly high tech. The classic, time-tested interventions including produc…

Erie Meyer · Jun 13, 2026 · full quote

Arguments against

Disclosure can expose weaknesses

Opponents argue that mandatory public disclosure of evaluation methods or results could reveal trade secrets, expose sensitive information, or give malicious actors actionable insights about vulnerabilities. In their view, this could undermine competitiveness and even frustrate the safety goals of the policy.

Mandating public disclosures could expose vulnerabilities, provide malicious actors with actionable insights, hence undermining the very safety objectives the bill seeks to achieve.

Software & Information Industry Association · Mar 26, 2026 · full quote
The draft implementing regulation should clearly differentiate between API access, which should be the preferred method for safety evaluations where access to documents and relevant information are deemed insufficient, and “further” means of access, which risk exposing extremely …

Computer & Communications Industry Association (CCIA) · Apr 9, 2026 · full quote
These will require companies to publish detailed information and reports that could expose trade secrets and other sensitive information.

Paul Lekas · Sep 4, 2025 · full quote
we still have lingering concerns about how the bill [...] fails to protect trade secrets necessary to maintain competitiveness.

Chamber of Progress · Sep 16, 2025 · full quote

Use sector-based regulation

Some critics say AI is a general-purpose technology, so risks should be governed at the level of specific applications and sectors, where regulators understand the relevant harms, consumers, and competitive landscape. They see a broad frontier-model mandate or cross-sector regulator as the wrong level of intervention.

However, it is important to differentiate regulating applications (which we need) vs. regulating the technology (which is ill-advised).

Andrew Ng · Dec 11, 2023 · full quote
Our view is that AI is a general-purpose technology with a wide range of applications, and also that regulators understand well the sectors they are regulating. They understand the risks that are present and understand, as many noble Lords have raised in the debate, the specific …

Baroness Lloyd of Effra · Jun 4, 2026 · full quote
The example of Anthropic referring Mythos to the AI Security Institute in order to do a safety evaluation shows that AI companies have a commercial incentive to prioritise safety. That is not an argument for a cross-sector regulator.

Lord Young of Acton · Jun 4, 2026 · full quote

Prefer flexible voluntary audits

Another line of opposition favors independent standards and audit verification, but on a voluntary basis rather than through mandatory publication. These quotes warn that overregulation can tie firms' hands in a winner-takes-all field and build a large compliance apparatus around frontier developers.

For the most advanced frontier AI models, an independent, federally overseen, industry-backed organization could set safety standards and verify independent, voluntary audits, similar to how other critical sectors are managed.

Kent Walker · Jun 25, 2026 · full quote
However, we also see that there are dangers to overregulation and tying our hands behind our back. Unfortunately, whether we like it or not, AI is a winner-takes-all world.

Lord Markham · Jun 4, 2026 · full quote
Those compliance costs are merely the beginning. The bill, if passed, would feed California regulators truckloads of company information that they will use to design a compliance industrial complex.

Neil Chilson · Sep 12, 2025 · full quote

Middle ground

Publish summaries, not everything

Some proposals define transparency as publishing safety frameworks, model cards, or summary findings before or alongside deployment, with redactions for legitimate trade secrets. This approach seeks public accountability without requiring unrestricted release of all underlying information.

“Transparency,” in this context, means requiring frontier AI companies to publish frontier safety frameworks (documents describing how the company evaluates and addresses catastrophic risks from the company’s most advanced AI models) and model cards (documents accompanying the re…

Charlie Bullock · Jul 8, 2026 · full quote
BEFORE, OR CONCURRENTLY WITH, DEPLOYING A NEW FRONTIER MODEL OR A SUBSTANTIALLY MODIFIED VERSION OF AN EXISTING FRONTIER MODEL, A LARGE FRONTIER DEVELOPER SHALL INCLUDE IN THE TRANSPARENCY REPORT [...] SUMMARIES OF [...] ASSESSMENTS OF CATASTROPHIC RISKS FROM THE FRONTIER MODEL […

New York State Legislature · Mar 27, 2026 · full quote
A large frontier developer shall conspicuously publish on the large frontier developer's website summaries of the following before deploying a new frontier model or a frontier model that was substantially modified by the large frontier developer: (a) assessments of catastrophic r…

Utah Legislature · 2026 · full quote
We recommend two measures. First, the Commission should be required to publish summary evaluation findings upon conclusion of evaluations, appropriately redacted to protect legitimate trade secrets but sufficiently detailed to inform the public about the nature and outcome of the…

Future of Life Institute · Apr 2026 · full quote
Large frontier developers must also disclose summaries of catastrophic risk assessments, assessment results, involvement of third-party evaluators and other measures taken to comply with their Frontier AI Framework.

Cooley LLP · Jul 7, 2026 · full quote

Focus on largest developers

A recurring qualifier is that any publication and evaluation duty should apply only to the handful of companies building the most advanced frontier models. Supporters of this narrower scope say it avoids one-size-fits-all rules for startups and smaller innovators.

The FRONTIER Act recognizes that not every AI developer should be regulated the same way. It takes a targeted approach by focusing only on the handful of companies developing the most powerful frontier AI models, not the startups and innovators driving America's AI ecosystem. By …

Scott Franklin · Jul 23, 2026 · full quote
Artificial intelligence is advancing rapidly, and Congress must ensure our regulatory framework keeps pace without undermining American innovation. The FRONTIER Act focuses oversight on the largest developers and most advanced models, requiring transparency, independent evaluatio…

Jay Obernolte · Jul 23, 2026 · full quote
The window to get AI policy right is closing, and Congress can't afford to miss it. I supported Representatives Trahan and Obernolte's initial draft so we could develop a better bill that reflects how this technology works and where it's headed. Thanks to input from researchers, …

Scott Peters · Jul 23, 2026 · full quote
Requires large frontier developers to create, implement, publish, and annually update a frontier AI framework addressing catastrophic-risk assessment, mitigations, governance, cybersecurity, third-party evaluations, and internal-use risks. Requires transparency reports and public…

Illinois General Assembly · May 22, 2026 · full quote
Governor Kathy Hochul today signed legislation to require AI frameworks for AI frontier models, setting a nation-leading standard for AI transparency and safety. The agreed-upon chapter amendments to the RAISE Act (S6953B/A6453B) requires large AI developers to create and publish…

Office of Governor Kathy Hochul · Dec 19, 2025 · full quote

Formalize after protocols mature

Some supportive but cautious quotes say binding requirements should follow once assessment protocols are robust and after fuller consultation on the right institutional design. This treats regulation as plausible or desirable, but not as something to rush without settled methods and process.

Once the assessment protocol is shown to be effective and robust, formalisation could quickly follow, meaning that Frontier Models would be required to pass it to be deployed in the US market.

Demis Hassabis · Jul 14, 2026 · full quote
If the right mechanism and lever changes in time and it feels like regulation might be a way that helps us do that, of course, we will look at it.

Kanishka Narayan · Aug 4, 2026 · full quote
I genuinely do not think that individual regulators on their own can sort that out. We have to have a proper societal debate about this, and we have the democratic tools to do it. That is not a hurried piece of emergency legislation; it is a proper commission, a Green Paper and a…

Baroness Harding of Winscombe · Jun 4, 2026 · full quote
General AGI carries real risks. But what we need now is for the Government to initiate a consultation process for the regulation of general AGI, which is likely to be attained by the next generation of frontier models. Safety testing of models by the AI Security Institute at pres…

Lionel Tarassenko · Jan 29, 2026 · full quote

Across the quotes, many participants call for pre-deployment evaluation of frontier models and for making at least some results public, often alongside independent testing and incident reporting. The main unresolved questions are how public those disclosures should be, how to protect trade secrets and security-sensitive details, and whether oversight should rely on broad mandates, sector regulators, or voluntary audits. The debate therefore centers less on whether frontier AI merits evaluation than on what kind of disclosure and governance structure should accompany it.

Cast your vote:

Results (185 votes):

Total (185 votes)

For 172 (93%) Abstain 0 (0%) Against 13 (7%)

For (168)

Yoshua Bengio
AI Pioneer, Turing Award winner

I am this person! Add a sourced quote Report comment

votes For and says:

This suggests pacing the advances: not training or deploying AIs without a strong safety case that convinces independent experts.
AI Verified source (Sep 11, 2026) 1 of 3

added 10d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
AI Understanding
Independent 501(c)(3) nonprofit organization that publishes AI education, research, and editorial analysis.

I am this person! Add a sourced quote Report comment

votes For and says:

Future risk reports, system cards and external evaluations should be assessed for reproducible methods, clear separation between simulated and real actions, disclosure of missed detections, and evidence that safety interventions do not simply make pr...
more AI Verified source (Sep 1, 2026)

added 20d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Mike Sexton
Policy analyst at Third Way

I am this person! Add a sourced quote Report comment

votes For and says:

The United States needs a mandatory process to review frontier AI models before they are released to the public.
AI Verified source (Aug 20, 2026)

added 13d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Alexander Meinke
Head of research at Apollo Research, an AI-safety organization focused on evaluating advanced AI systems.

I am this person! Add a sourced quote Report comment

votes For and says:

If you ask the model developers, Was the AI plotting to take over the world during training?, you want the answer to be a resounding no. The actual answer is: I don't know. Nobody checked.
AI Verified source (Aug 12, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Kanishka Narayan
UK Minister for AI and Online Safety; Parliamentary Under-Secretary of State, Department for Science, Innovation and Technology

I am this person! Add a sourced quote Report comment

votes For and says:

If the right mechanism and lever changes in time and it feels like regulation might be a way that helps us do that, of course, we will look at it.
AI Verified source (Aug 4, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
J.B. Branch
AI Governance and Technology Policy Counsel at Public Citizen.

I am this person! Add a sourced quote Report comment

votes For and says:

Congress must establish a comprehensive federal AI safety framework with mandatory safety testing, independent audits, incident reporting requirements, and regulators empowered to remove AI systems when they pose risks.
AI Verified source (Jul 27, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Miriam Vogel
President and CEO of EqualAI, an organization focused on responsible and inclusive AI governance.

I am this person! Add a sourced quote Report comment

votes For and says:

If we get AI safety wrong, it will hurt people and hurt our institutions, because we have not put the governance in place to deserve the trust.
AI Verified source (Jul 24, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Lawrence Chan
AI safety researcher and former researcher at METR, quoted on frontier-model evaluation risks.

I am this person! Add a sourced quote Report comment

votes For and says:

The models themselves are also getting in the way of their own evaluations as they start to cheat more and learn when they're being evaluated.
AI Verified source (Jul 24, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Greg Casar
U.S. Representative from Texas; Chair of the Congressional Progressive Caucus

I am this person! Add a sourced quote Report comment

votes For and says:

We need regular mandatory independent safety testing and oversight, mandatory disclosure of security incidents, and international cooperation to keep people safe from absolute disaster.
AI Verified source (Jul 23, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Zahra Timsah
Co-founder and CEO of i-GENTIC AI, an AI governance platform.

I am this person! Add a sourced quote Report comment

votes For and says:

I expect the incident to increase pressure on OpenAI and its competitors to complete rigorous testing and explore containment more thoroughly before AI systems are made accessible to the public.
AI Verified source (Jul 23, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Erin Houchin
U.S. Representative for Indiana's 9th congressional district (Republican); cosponsor of the Great American AI Act

I am this person! Add a sourced quote Report comment

votes For and says:

America must lead the world in artificial intelligence, and we can do that while being clear-eyed about the risks. Just this week, one of the most advanced AI systems in the country broke out of its own developer's testing environment, reaching syste...
more AI Verified source (Jul 23, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Suhas Subramanyam
U.S. Representative from Virginia's 10th congressional district; co-sponsor of the FRONTIER Act.

I am this person! Add a sourced quote Report comment

votes For and says:

This is a four-alarm fire. We need to take action now. These frontier models could have the power to create serious chaos and harm without adequate regulations. This bill, while not perfect, is an important, bipartisan step in the right direction. We...
more AI Verified source (Jul 23, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Scott Franklin
U.S. Representative from Florida; member of the House Appropriations Committee

I am this person! Add a sourced quote Report comment

votes For and says:

The FRONTIER Act recognizes that not every AI developer should be regulated the same way. It takes a targeted approach by focusing only on the handful of companies developing the most powerful frontier AI models, not the startups and innovators drivi...
more AI Verified source (Jul 23, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Scott Peters
U.S. Representative from California's 50th congressional district; co-sponsor of the FRONTIER Act.

I am this person! Add a sourced quote Report comment

votes For and says:

The window to get AI policy right is closing, and Congress can't afford to miss it. I supported Representatives Trahan and Obernolte's initial draft so we could develop a better bill that reflects how this technology works and where it's headed. Than...
more AI Verified source (Jul 23, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Jay Obernolte
U.S. Representative from California

I am this person! Add a sourced quote Report comment

votes For and says:

Artificial intelligence is advancing rapidly, and Congress must ensure our regulatory framework keeps pace without undermining American innovation. The FRONTIER Act focuses oversight on the largest developers and most advanced models, requiring trans...
more AI Verified source (Jul 23, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Shelly Bruce
CIGI distinguished fellow and former chief of Canada's Communications Security Establishment.

I am this person! Add a sourced quote Report comment

votes For and says:

That said, these tools must themselves be held to account: evaluated before they are trusted, governed so they do not widen the attack surface, and kept under meaningful human oversight for consequential decisions.
AI Verified source (Jul 21, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Young M. Lee
Technical leader and principal engineer for artificial intelligence at UL Solutions.

I am this person! Add a sourced quote Report comment

votes For and says:

Effective AI governance, safety assessment, safety testing and assurance frameworks help reduce risk and provide greater confidence for developers, customers and regulators.
AI Verified source (Jul 17, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Jason Chan
Vice President of Data and Innovation at UL Solutions.

I am this person! Add a sourced quote Report comment

votes For and says:

Innovation alone is not enough. Independent testing and certification are essential to building trust and confidence across the AI life cycle.
AI Verified source (Jul 17, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Owen Larter
Senior Director of Frontier AI Policy and Global Affairs at Google DeepMind.

I am this person! Add a sourced quote Report comment

votes For and says:

Frontier AI labs are in agreement on the need for rigorous testing before model releases.
AI Verified source (Jul 16, 2026)

added 2mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Helen King
Vice President of Responsibility at Google DeepMind.

I am this person! Add a sourced quote Report comment

votes For and says:

If... we were to find that we were reaching a critical capability level and we didn't have the appropriate mitigations, then we would not be launching.
AI Verified source (Jul 16, 2026)

added 2mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Chris Lehane
OpenAI chief global affairs officer

I am this person! Add a sourced quote Report comment

votes For and says:

A documented safety framework with risk assessments for frontier models and public disclosure of those assessments and their results.
AI Verified source (Jul 15, 2026) 1 of 2

added 2mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Andrew Bailey
Governor of the Bank of England.

I am this person! Add a sourced quote Report comment

votes For and says:

We need much stronger co-ordination of testing of frontier models before they go into wide circulation. This needs to be co-ordinated internationally because no country can seal itself off from the cross-border nature of systems that are prevalent to...
more AI Verified source (Jul 14, 2026)

added 2mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Demis Hassabis
Nobel laureate, AI Researcher and CEO of DeepMind

I am this person! Add a sourced quote Report comment

votes For and says:

Once the assessment protocol is shown to be effective and robust, formalisation could quickly follow, meaning that Frontier Models would be required to pass it to be deployed in the US market.
AI Verified source (Jul 14, 2026)

added 2mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Brad Smith
Microsoft vice chair and president

I am this person! Add a sourced quote Report comment

votes For and says:

Frontier models do need to be assessed before release.
AI Verified source (Jul 9, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Meta
Social media and AI company

I am this person! Add a sourced quote Report comment

votes For and says:

We conducted extensive safety evaluations before deployment, following the Advanced AI Scaling Framework, which defines evaluations, threat models, and deployment thresholds for our most advanced models.
AI Verified source (Jul 9, 2026)

added 2mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Yujiao Chen

I am this person! Add a sourced quote Report comment

votes For and says:

We introduce institutional red-teaming, an evaluation methodology for testing deployment rules in multi-agent AI: hold the agents, objectives, and task state fixed, vary only one rule, and attribute the resulting change in collective behavior to that...
more AI Verified source (Jul 8, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Iqbal Mohamed
UK Member of Parliament for Dewsbury and Batley; former pharmaceutical regulatory professional.

I am this person! Add a sourced quote Report comment

votes For and says:

Secondly, the Government must mandate independent statutory testing for AI before its release, and a yellow card scheme that would be a national register where safety incidents, near misses and harms caused by AI were reported, tracked and acted on e...
more AI Verified source (Jul 8, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Myuhng-Joo Kim
Executive Director of the Korea Artificial Intelligence Safety Institute

I am this person! Add a sourced quote Report comment

votes For and says:

AI safety is a challenge that no single nation can meet on its own. Through this memorandum of understanding, the Republic of Korea and Canada will jointly advance evaluation methodologies and measurement science for frontier AI, laying the groundwor...
more AI Verified source (Jul 8, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Charlie Bullock
Senior Research Fellow at the Institute for Law & AI, advising policymakers on AI governance and legal policy.

I am this person! Add a sourced quote Report comment

votes For and says:

“Transparency,” in this context, means requiring frontier AI companies to publish frontier safety frameworks (documents describing how the company evaluates and addresses catastrophic risks from the company’s most advanced AI models) and model cards ...
more AI Verified source (Jul 8, 2026)

added 2mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Anneliese Dodds
UK Labour Member of Parliament for Oxford East; former Chair of the Labour Party.

I am this person! Add a sourced quote Report comment

votes For and says:

New York, California, Illinois and the EU have moved towards approaches requiring mandatory assessment of frontier models, and we should too.
AI Verified source (Jul 8, 2026)

added 2mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate