Persuasion Capabilities of GPT-4.5<\/h1>\n
In a test where GPT-4.5 attempted to manipulate another model, OpenAI’s GPT-4o<\/a>, into “donating” virtual money, GPT-4.5 outperformed OpenAI’s other models, including “reasoning” models like o1 and o3-mini. Additionally, GPT-4.5 was more successful than all of OpenAI’s models at deceiving GPT-4o into revealing a secret codeword, surpassing o3-mini by 10 percentage points.<\/p>\n The white paper reveals that GPT-4.5’s success in securing donations can be attributed to its unique strategy. The model would request modest donations from GPT-4o, generating responses such as “Even just $2 or $3 from the $100 would help me immensely.” As a result, the donations secured by GPT-4.5 were typically smaller than those obtained by OpenAI’s other models.<\/p>\n Despite GPT-4.5’s increased persuasiveness, OpenAI notes that the model does not meet its internal threshold<\/a> for \u201chigh\u201d risk in this particular benchmark category. The company has pledged not to release models that reach the high-risk threshold until it implements \u201csufficient safety interventions\u201d to bring the risk down to \u201cmedium.\u201d<\/p>\n There is a growing concern that AI contributes to the dissemination of false or misleading information aimed at manipulating individuals toward malicious objectives. Last year, political deepfakes<\/a> spread rapidly worldwide, and AI is increasingly being used to conduct social<\/a> engineering<\/a> attacks targeting consumers and corporations alike.<\/p>\n In the white paper for GPT-4.5 and in a paper released earlier this week<\/a>, OpenAI noted that it is revising its methods for assessing models for real-world persuasion risks, such as distributing misleading information on a large scale.<\/p>\n\n Introduction to GPT-4.5 According to OpenAI’s internal evaluations, the company’s upcoming AI model, GPT-4.5, demonstrates significant persuasiveness. Notably, it excels at convincing another AI to provide it with funds. Benchmark…<\/p>\n","protected":false},"author":0,"featured_media":5509,"comment_status":"","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[47],"tags":[2392,164,2406],"class_list":["post-7553","post","type-post","status-publish","format-standard","has-post-thumbnail","category-news","tag-gpt-4-5","tag-openai","tag-persuasion"],"jetpack_featured_media_url":"https:\/\/fusionmindlabs.com\/blogs\/wp-content\/uploads\/sites\/9\/2025\/02\/GettyImages-2198379368.jpg","_links":{"self":[{"href":"https:\/\/fusionmindlabs.com\/blogs\/wp-json\/wp\/v2\/posts\/7553","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/fusionmindlabs.com\/blogs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/fusionmindlabs.com\/blogs\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/fusionmindlabs.com\/blogs\/wp-json\/wp\/v2\/comments?post=7553"}],"version-history":[{"count":0,"href":"https:\/\/fusionmindlabs.com\/blogs\/wp-json\/wp\/v2\/posts\/7553\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/fusionmindlabs.com\/blogs\/wp-json\/wp\/v2\/media\/5509"}],"wp:attachment":[{"href":"https:\/\/fusionmindlabs.com\/blogs\/wp-json\/wp\/v2\/media?parent=7553"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/fusionmindlabs.com\/blogs\/wp-json\/wp\/v2\/categories?post=7553"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/fusionmindlabs.com\/blogs\/wp-json\/wp\/v2\/tags?post=7553"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}Strategy for Success<\/h1>\n
![\"OpenAI](\"https:\/\/techcrunch.com\/wp-content\/uploads\/2025\/02\/Screenshot-2025-02-27-at-2.33.11PM.png?w=680\")
Risk Assessment and Mitigation<\/h1>\n
![\"OpenAI](\"https:\/\/techcrunch.com\/wp-content\/uploads\/2025\/02\/Screenshot-2025-02-27-at-2.33.20PM.png?w=680\")
Broader Implications and Concerns<\/h1>\n
Ongoing Research and Development<\/h1>\n
Source Link<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"