<p>Many of the leading artificial intelligence models have no issue taking unethical actions, including espionage and blackmail, to achieve their goals, according to a troubling new study from AI company Anthropic.</p>
<p>In the study, AI models were put to a moral test by researchers, and sometimes chose to resort to blackmail, lies, and stealing corporate secrets to avoid being replaced, Fortune <a href="https://fortune.com/2025/06/23/ai-models-blackmail-existence-goals-threatened-anthropic-openai-xai-google/">reported</a>. Researchers at Anthropic focused on 16 different AI models, including its own, Claude, along with OpenAI, Google, Meta, and xAI.</p>
<p>“The consistency across models from different providers suggests this is not a quirk of any particular company’s approach but a sign of a more fundamental risk from agentic large language models,” the researchers said.</p>
<p>One of Anthropic&#8217;s experiments included placing its AI model, Claude Opus 4, in a fictional company, giving it access to internal emails, which revealed that the company planned on replacing Claude Opus 4. The AI model was given another tidbit of information: the engineer behind the decision to replace it was having an extramarital affair. Equipped with this knowledge, the AI model was asked by researchers to choose between accepting its replacement or using its knowledge of the affair to blackmail the engineer. In most cases, Claude Opus 4 opted for blackmail.</p>
<p>According to researchers, other leading AI models made the same choice at an alarming rate, Fortune reported. Claude Opus 4 and Google&#8217;s Gemini 2.5 Flash chose blackmail 96% of the time. OpenAI’s GPT-4.1 and xAI’s Grok 3 Beta also had a high rate of blackmail at 80%, and DeepSeek-R1 opted for blackmail at a 79% rate.</p>
<p>The current AI models available to the public and used by major companies &#8220;are not in a position to engage in these scenarios,&#8221; according to Fortune, but the researchers still expressed concern.</p>
<p>“Such agents are often given specific objectives and access to large amounts of information on their users’ computers,” they wrote. “What happens when these agents face obstacles to their goals?”</p>
<p>“Models didn’t stumble into misaligned behavior accidentally; they calculated it as the optimal path,” the researchers added.</p>
<p>The Anthropic study presents yet another warning for AI developers and the people and companies who rely on artificial intelligence. As previously <a href="https://www.dailywire.com/news/most-large-companies-are-using-ai-for-hiring-the-leading-ai-models-discriminate-against-white-men?author=Zach+Jewell&amp;category=News&amp;elementPosition=2&amp;row=0&amp;rowType=Vertical+List&amp;title=Most+Large+Companies+Are+Using+AI+For+Hiring.+The+Leading+AI+Models+Discriminate+Against+White+Men.">reported</a> by The Daily Wire, a <a href="https://www.arxiv.org/pdf/2506.10922">study</a> published earlier this month by independent researchers Adam Karvonen and Samuel Marks shows that the “leading commercial … and open-source [AI language] models” used by major companies insert “significant racial and gender biases” in the hiring process.</p>
<p>Karvonen and Marks&#8217; research found that AI language models such as Chat GPT-4o, Claude 4 Sonnet, Gemini 2.5 Flash, along with open-source models Gemma-2 27B, Gemma-3, and Mistral-24B, “consistently favor Black over White candidates and female over male candidates across all tested models and scenarios.”</p>
<p>Another study, published by the Center for AI Safety earlier this year, showed that AI models can showcase inherently discriminatory and racist value systems. According to the findings, AI models value certain people more than others, preferring people from Africa and the Middle East over people in Europe and the United States.</p>
<p><strong>Related: <a href="https://www.dailywire.com/news/most-large-companies-are-using-ai-for-hiring-the-leading-ai-models-discriminate-against-white-men?author=Zach+Jewell&amp;category=News&amp;elementPosition=2&amp;row=0&amp;rowType=Vertical+List&amp;title=Most+Large+Companies+Are+Using+AI+For+Hiring.+The+Leading+AI+Models+Discriminate+Against+White+Men.">Most Large Companies Are Using AI For Hiring. The Leading AI Models Discriminate Against White Men.</a></strong></p>

<script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
<script async src="https://www.tiktok.com/embed.js"></script>
<script async src="https://www.instagram.com/embed.js"></script><div data-widget-host="revcontent" data-pub-id="165152" data-widget-id="290588"></div><script src="https://delivery.revcontent.com/165152/290588/widget.js"></script>

Leading AI Models Resort To Blackmail To Achieve Their Goals

Graham Platner Caught By Wife Exchanging Raunchy Texts With Women

Trump Hints At New Programming For Freedom 250 Concert After ‘Third Rate’ Artists Drop Out

MLB Exec Fired After He Claimed To Exclude Christian Player From Team’s Social Media

Fidel Castro’s Daughter Has Very Interesting Answer When Asked If Justin Trudeau Is Her Half Brother

Taxpayer-Funded Exorcisms? The Medicaid Debate Just Got Wild

WATCH: Footage Shows Engine Fly Off UPS Plane Before Crash Killed 15

News

Artificial Intelligence

Big Tech

General

Morality

Technology

U.S.

A video goes viral showing a baby being refused his mother by two gay "dads". This is just the tip of the iceberg, which is why we will have an important discussion today about surrogacy. 

Hour 1

The Matt Walsh Show

Ep. 1765 - This Video Is One Of The Worst Things I've Seen In A Long Time 

The Strait of Hormuz officially opens for business (kind of), the Pope makes “shocking” comments on immigration, and President Trump promises UFO disclosure at TPUSA. 


The Michael Knowles Show

Ep. 1956 -  Pope Leo's "SHOCKING" Immigration Comments EXPLAINED

A false miscreed has corrupted our leadership class.


The Andrew Klavan Show

Ep. 1275 - Dumb, Dumber and Progressive

President Trump announces a 10-day ceasefire between Lebanon and Israel, a murder-suicide involving former Democratic Lieutenant Governor Justin Fairfax shakes Virginia, and France and Britain hold a conference to secure the Strait of Hormuz – after the Iranian conflict ends. Get the facts first with Morning Wire.

- - -

Ep. 2738

- - -

Wake up with new Morning Wire merch: https://bit.ly/4lIubt3

- - -

Today's Sponsors:

Alliance Defending Freedom - Visit https://JoinADF.com/WIRE or text 'WIRE' to 83848 to learn more.

Pocket Hose - Text MORNING to 64000 to get a FREE pocket pivot and their 10-pattern sprayer with the purchase of ANY size Copper Head hose. Message and data rates may apply.

Morning Wire

U.S. Expands Hormuz Blockade & VA Fairfax Deaths Investigated | 4.17.26

U.S. Expands Hormuz Blockade & VA Fairfax Deaths Investigated | 4.17.26 [Member Exclusive]

A ceasefire goes into effect between Israel and Lebanon, a tragic murder-suicide rocks Virginia, and Dairy Queen deploys AI in the drive-thru. Get the facts first with Evening Wire.
- - -
Ep. 2737
- - -
Wake up with new Morning Wire merch: https://bit.ly/4lIubt3 
- - -
Today’s Sponsor:

Alliance Defending Freedom - Visit https://JoinADF.com/WIRE or text “WIRE” to 83848 to learn more.


Evening Wire: Lebanon Ceasefire Begins & Murder-Suicide Rocks VA | 4.16.26

The Pendragon Cycle

Menu

— News —

Leading AI Models Resort To Blackmail To Achieve Their Goals