OpenAI updated its safety framework—but no longer sees mass manipulation and disinformation as a critical risk

Quick Summary:

OpenAI’s updated AI safety framework drops key pre-release testing requirements—including for persuasive or manipulative capabilities—sparking concern among experts.

That framework details how the company monitors the AI models it is building for potentially catastrophic  dangers—everything from the possibility the models will help someone create a biological weapon to  their ability to assist hackers to the possibility that the models will self-improve and escape human control.

Oren Etzioni, former CEO of the Allen Institute for AI and founder of TrueMedia, which offers tools to fight AI-manipulated content, also expressed concern.“Downgrading deception strikes me as a mistake given the increasing persuasive power of LLMs,” he said in an email.“One has to wonder whether OpenAI is simply focused on chasing revenues with minimal regard for societal impact.

---

More Pictures

---

Read Quick SummarySave Time

or

Article Details

Author / Journalist: Sharon Goldman, Jeremy Kahn

Category: EconomyBusiness

Markets: United States

Topics: OpenAI A.I.

Source Website Secure: Yes (HTTPS)

News Sentiment: Neutral

Fact Checked: Legitimate

Article Type: News Report

Published On: 2025-04-16 @ 19:09:31 (5 days ago)

News Timezone: GMT +8:00

News Source URL: fortune.com

Language: English

Article Length: 1076 words

Reading Time: 6 minutes read

Sentences: 4 lines

Sentence Length: 269 words per sentence (average)

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

Copyright Owner: © Fortune

News ID: 27954803

View Article Analysis

About Fortune

Main Topics: EconomyBusiness

Official Website: fortune.com

Update Frequency: 39 posts per day

Year Established: 1929

Headquarters: United States

News Last Updated: 8 hours 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: #95

Publisher Details

Frequently Asked Questions