Opinion

The summit underscored the deep divisions in how different stakeholders view AI regulation.

Does a rose by any other name smell as sweet? 

A comprehensive analysis of current AI agents reveals a significant lack of information about safety policies and evaluations.

Meta's framework is a good start, but some glaring issues need improvement.

AI-themed advertisements are the first step toward a severe risk of “gradual disempowerment” for humanity.

The only way to make AI benchmarks even harder is to have them written and scored by AI.

AI reasoning systems are capable of formulating complex internal narratives that diverge significantly from their outward responses.

The less people know about AI, the more likely they are to embrace it.

CAIP will be looking to hear 5 questions asked at Lutnick's confirmation hearing.

AI is a critical topic for any U.S. president to address.

If it's important enough to spend $500 billion on, it’s important enough to have a specific destination in mind.

Broad Congressional support is needed for high-quality AI governance

A new paper on Best-of-N Jailbreaking demonstrates the inherent unreliability of reinforcement-based alignment

After two years of congressional deliberation on artificial intelligence, we need more than careful analysis

The continued lack of a legislative safety net for AI’s unintended consequences puts us all at risk

An op-ed in Tech Policy Press on AI's likely effects on musicians and how to guide them in a better direction.

The United Healthcare tragedy reminds us that critical health coverage decisions cannot be safely delegated to inscrutable AI algorithms

We have hard evidence that AI is lying, scheming, and protecting itself, but developers don’t care

There’s more science to be done, but it’s not too early to start collecting reports from AI developers

The most recent CAIP podcast explores four principles to address ever-evolving AI

The United States hosted the inaugural meeting of a growing global network

Analyzing present and future military uses of AI

Slower AI progress would still move fast enough to radically disrupt American society, culture, and business

AISI conducted pre-deployment evaluations of Anthropic's Claude 3.5 Sonnet model

Op-ed in Tech Policy Press on AI safety testing and US-China competition

For the US to lead on AI, AISI needs resources and formal authorization from Congress

Countless American sports fans are suffering emotional despair and financial ruin as a result of AI-powered online sports betting

5 horror movie tropes that explain trends in AI

The FTC enacted its final rule banning fake AI reviews

Reclaim safety as the focus of international conversations on AI

The jury is out on the social and health effects of these tools

Unfortunately, not all AI agent applications are low-risk

Why is there still inaction in Congress?

Many voters prefer a careful, safety-focused approach to AI development over a strategy emphasizing speed and geopolitical competition

States cannot afford to wait any longer to begin setting AI safety rules

Creating a plan, anticipating challenges, and executing a coordinated response saves lives and protects communities

No one man, woman, or machine should have this much power over the future of AI

It’s time for Congress to act

The Senate Select Committee on Intelligence bipartisanly, publicly, and loudly implored the American people and the private sector to remain vigilant against election interference

CAIP hopes that Walz and Vance will tell their fellow Americans where they stand on AI safety legislation

OpenAI's lobbying has expanded, and it's crowding out dialogue on safety.

If Governor Newsom cares about AI safety, he'll sign SB 1047

The US is punching below its weight when it comes to funding its AI Safety Institute (AISI)

CAIP traveled to New York City to hear what local AI professionals have to say about AI's risks and rewards

Who is vouching for the safety of OpenAI’s most advanced AI system?

Engineers continue discovering techniques that boost AI performance after the main training phase

A new polls shows once again that the American public is profoundly skeptical of AI and worried about its risks

America’s best-beloved circles up a crew of technologists, humanists, and a law enforcer on what’s next for humanity in AI

Out of 30 campaign websites reviewed, only 4 had even a single clear position on AI policy

The Center for AI Policy (CAIP) believes the 2024 Presidential candidates need to take a stand on AI safety

A 10-year-old girl accidentally hanged herself while trying to replicate a “Blackout Challenge” shown to her by TikTok’s video feed.

Yet another example why we need safe and trustworthy AI models.

Political campaigns should disclose when they use AI-generated content on radio and television.

Last Monday night (8/19/24), the Democratic Party approved its 2024 Party Platform. The platform’s general rhetoric hits all the key themes of AI safety.

Even if we plug the holes in our porous firewalls, there’s another problem we have to solve in order to win an AI arms race: alignment.

Reflections on a trip to DEFCON 2024

Human-quality speech presents a heightened risk that AI is used for fraud, misinformation, and manipulation

Why government oversight must complement corporate commitments

CAIP proposes that AI companies report their cybersecurity protocols against a set of key metrics

This step forward doesn’t mean that AI is free from sexual exploitation

Meta essentially ran a closed-source safety check on an open-source AI system

This highlights challenges in anticipating malicious uses of AI 

Today's CrowdStrike-Microsoft outage is case in point

To promote human flourishing, AI tools must be safe

As time goes on, the ways in which AI can enhance itself will multiply

AI safety and responsibility are core themes of NATO's AI Strategy

Regardless of the outcome, OpenAI needs stronger whistleblower protections

The security breach at OpenAI should raise serious concerns among policymakers

Like the links on the second page of Google’s search results, these principles are something of a mixed bag

Corporate incentives do not ensure optimal outcomes for public safety

Neil Chilson's recent critique of the Center for AI Policy's model AI safety legislation is deeply misleading

Both Biden and Trump agree that AI is scary, but what do they plan to do about these dangers?

An op-ed in the Capital Journal

Qualified third parties should audit AI systems and verify their compliance with federal laws and regulations

We hope that at their next product launch, Apple will address AI safety

Aschenbrenner argues that AI systems will improve rapidly

The responsible thing to do is to take their warnings seriously

Our views on the latest AI resources from NIST

Unpacking the details of Microsoft's latest announcement about expanding its responsible AI team

Right now, only Big Tech gets to decide whether AI systems are safe

CAIP's Executive Director participated in a panel discussion hosted by the Social Security Administration

Genius-level AI will represent a total paradigm shift

March 2024 Op-Ed in The Hill: Robocalls Are the Least of Our AI Worries

Protecting individual autonomy and society from potential harms

Our Executive Director's testimony in support of HB1062

Our thoughts on AI topics pressing for global security to discuss this MSC

Why America needs AI legislation