OpenAI is set to unveil a multimodal AI model that can communicate with users and recognize objects, per reports. Anonymous sources said this reveal might happen on Monday.

The new model can analyze pictures and audio faster and more accurately than current transcription and text-to-speech technologies, according to a report from The Verge. It is anticipated to improve customer service by helping operators understand caller tone and sarcasm. According to the model, it may also help pupils with algebra and translation signs.

Although said to exceed the GPT-4 Turbo in some question-answering tasks, sources warn that the model is not error-proof.

Based on developer Ananay Arora's findings, OpenAI may be adding ChatGPT functionality for phone conversations. This theory is supported by evidence of real-time audio and video server provisioning.

Amid these speculations, tech fans anticipate the upcoming OpenAI announcement to be live-streamed on Monday at 10 a.m. PT or 1 p.m. ET.

Sam Altman Addresses OpenAI Search Engine Rumors

In a separate development, Sam Altman, CEO of OpenAI, denied speculations of a search engine launch before Google I/O. Altman disputes recent allegations that OpenAI will release a search product before Google's big event.

Altman said an OpenAI statement on Monday morning does not involve GPT-5 or any search engine. He hinted at something unexpected and extraordinary. OpenAI has only mentioned ChatGPT and GPT-4 improvements.

According to reports, OpenAI is looking for Google employees to form a specialized team to accelerate product development. Bloomberg has reported OpenAI's search product development, but a debut date is unknown.

Read Also: Meta Unveils 3 New Tiers to Expand Paid Verification Service for Businesses

OpenAI Considers to Allow AI Generated Porn?

Adding to these speculations, the ChatGPT maker is reportedly considering letting users create pornography and other not safe-for-work (NSFW) content.

OpenAI considers allowing developers and users to generate age-appropriate explicit material in a Wednesday document explaining its AI development protocols.

Joanne Jang, who drafted the letter, said the tech company will debate sexual language and naked photos on its platforms. However, she stressed that deepfakes would remain prohibited. She emphasized user control, legal and ethical compliance, and open discourse, as well as understanding pornography's subjectivity.

While prohibiting deepfakes with adult content, OpenAI is studying how its technology may help produce responsible adult content, including erotica, graphic violence, foul language, and explicit material.

An OpenAI clarified on Thursday that the artificial intelligence tech firm has no plans to create AI-generated pornography. The representative reiterated OpenAI's ban on deepfakes. The statement also stressed, "thoughtfully navigating discussions surrounding sexuality within appropriate contexts for various age groups."

Recent months have seen criticism about AI-generated pornography, notably deepfake. AI-generated sexual photos quickly spread on social media in January, affecting Taylor Swift. Moreover, deepfake porn also impacted teenage students of Beverly Hills High School in California and Congresswoman Alexandria Ocasio-Cortez.

Related Article: Google Cloud Partners with CrowdStrike to Transform AI-Native Cybersecurity