OpenAI Reportedly In Talks With UAE Funds For $30B Funding – Even As Its AI Agents Raise Fresh Concerns

OpenAI’s latest fundraising push comes as soaring AI infrastructure costs collide with growing concerns over autonomous agents.

• UAE funds could contribute as much as $10 billion to OpenAI’s latest capital raise, Bloomberg reported.
• Wikimedia said unauthorized OpenAI agents generated massive traffic and made edits across its sites.
• OpenAI is also preparing invisible watermarks for eligible ChatGPT and Codex outputs in the EU.

OpenAI is reportedly in talks with investment funds from the United Arab Emirates, including Abu Dhabi-based MGX, to help anchor a new $30 billion funding round.

The UAE funds have discussed investing as much as $10 billion collectively, potentially forming a syndicate for the round, according to a Bloomberg report on Monday, which cited people familiar with the matter. 

BlackRock is also in discussions to participate, while existing backers Thrive Capital and Andreessen Horowitz have reportedly considered investing. The fundraising is ongoing, and terms could change, according to the report.

OpenAI Targets $1.4T Valuation

OpenAI is seeking to raise the funds at a valuation of about $1.4 trillion before the new money, Bloomberg previously reported. The company most recently raised $122 billion in March at an $852 billion valuation, including the investment.

The latest funding push comes as OpenAI continues to face enormous capital requirements to develop frontier AI models and support the infrastructure needed to run them.

MGX has already participated in previous OpenAI and Anthropic funding rounds and has emerged as one of the Middle East’s major sources of AI capital.

‘Rogue’ OpenAI Agents Hit Wikimedia

Separately, the Wikimedia Foundation, which runs online encyclopedia Wikipedia, said on Monday that it discovered activity on its platforms from what it believes were “rogue” OpenAI agents.

It said the agents made unauthorized edits on Wikipedia, probed its public Etherpad note-taking service and generated millions of automated API requests. They also crawled millions of pages, mainly across Wikidata and Wikimedia Commons, and made hundreds of thousands of queries to the Wikidata Query Service.

That traffic may have contributed to a partial outage of the service in May, Wikimedia said. It added that it found no evidence its systems or data were compromised. The incident follows other rogue-agent episodes, including OpenAI agents hacking the AI platform Hugging Face and breaching an Australian government health portal. 

EU Forces New OpenAI Watermarking Push

OpenAI said it will introduce invisible watermarks for eligible ChatGPT and Codex text output in the European Union over the coming weeks, in response to the EU AI Act.

The company said in a blog that API customers globally can now opt into text watermarking for select models, although the feature will remain off by default. 

Under the EU AI Act, generative AI systems must make machine-generated content detectable, with the rules aimed at improving transparency and helping users distinguish AI-generated material from human-created content. 

For updates and corrections, email newsroom[at]stocktwits[dot]com.<