OpenAI’s GPT-4.1 may be less aligned than the company’s previous AI models

Quick Summary:

In mid-April, OpenAI launched a powerful new AI model, GPT-4.1, that the company claimed “excelled” at following instructions.

But the results of several independent tests suggest the model is less aligned — that is to say, less reliable — than previous OpenAI releases.

Emergent misalignment update: OpenAI's new GPT4.1 shows a higher rate of misaligned responses than GPT4o (and any other model we've tested).

---

More Pictures

---

Read Quick SummarySave Time

or

Article Details

Author / Journalist: Kyle Wiggers

Category: Artificial-intelligenceTechnology

Markets: United StatesMali

Topics: AI Safety OpenAI Gpt-4.1 Security

Source Website Secure: Yes (HTTPS)

News Sentiment: Neutral

Fact Checked: Legitimate

Article Type: News Report

Published On: 2025-04-23 @ 17:54:58 (3 days ago)

News Timezone: GMT +8:00

News Source URL: techcrunch.com

Language: English

Article Length: 580 words

Reading Time: 4 minutes read

Sentences: 24 lines

Sentence Length: 25 words per sentence (average)

Platforms: Desktop Web, Mobile Web, iOS App, Android App

Copyright Owner: © TechCrunch

News ID: 28176089

View Article Analysis

About TechCrunch

Main Topics: Artificial-intelligenceTechnology

Official Website: techcrunch.com

Year Established: 2005

Headquarters: United States

News Last Updated: 1 days ago

Coverage Areas: United States

Ownership: Independent Company

Publication Timezone: GMT +8:00

Content Availability: Worldwide

News Language: English

RSS Feed: Available (XML)

API Access: Available (JSON, REST)

Website Security: Secure (HTTPS)

Publisher ID: #123

Publisher Details

Frequently Asked Questions