OpenAI is reconsidering the critical clause in its corporate structure that limits Microsoft’s access to its advanced AI systems upon achieving artificial general intelligence (AGI), reports the Financial Times.
Initially introduced to ensure ethical oversight, the provision mandates that control over AGI technologies transfers to OpenAI’s nonprofit board, preventing potential misuse for commercial gain. However, this clause now poses challenges to OpenAI’s partnership with Microsoft, which is integral to its technological and financial ecosystem.
Microsoft, a key stakeholder in OpenAI with investments totaling nearly $14 billion since 2019, has embedded OpenAI’s models into its Azure cloud platform and enterprise tools like Microsoft 365 Copilot.
Revising the AGI clause would allow Microsoft to retain access to OpenAI’s future technologies, solidifying a partnership that has driven significant advancements in generative AI.
According to OpenAI, AGI refers to “highly autonomous systems that outperform humans at most economically valuable work” and is explicitly excluded from commercial licensing agreements under the current structure.
During The New York Times’ DealBook Summit, CEO Sam Altman highlighted the importance of aligning OpenAI’s strategic priorities with its investor relationships, stating, “If someone invests in a competitor, they don’t get access to our roadmap.”
ChatGPT Pro Expands Enterprise Offerings
In tandem with discussions around the AGI clause, OpenAI launched ChatGPT Pro on December 5, a premium subscription tier priced at $200 per month. Targeted at professionals and enterprise users, ChatGPT Pro offers unlimited access to OpenAI’s most advanced models, including GPT-4o and the enhanced o1 reasoning system.
Related: ChatGPT’s o1 Pro Mode Falls Short in SimpleBench, Will GPT-4.5 Turn the Tide?
First introduced in preview in September, the o1 model has been refined to meet the demands of high-performance users. The Pro tier features o1 Pro Mode, which leverages additional computational power to tackle complex challenges, delivering superior accuracy and reliability in fields like coding, competition mathematics, and scientific analysis.
OpenAI describes the o1 Pro Mode as “designed to provide more reliable, accurate, and comprehensive responses” across advanced use cases. The plan also includes Advanced Voice Mode, enabling conversational interactions for professional applications, and the Canvas Interface, a developer-focused tool that allows real-time editing of AI-generated code.
Financial Pressures Shape OpenAI’s Strategic Decisions
Despite its impressive estimated valuation of $157 billion, OpenAI faces mounting financial pressures. The company projects a $5 billion loss in 2024, with cumulative losses expected to reach $44 billion by 2028. Much of this financial strain stems from the high costs of training and deploying large-scale AI models, with annual compute expenses anticipated to climb to $9.5 billion by 2026.
To somehow address these challenges, OpenAI has partnered with semiconductor manufacturers TSMC and Broadcom to develop custom AI chips. These chips, slated for production in 2026, are set to utilize TSMC’s A16 node technology for efficient model training and Broadcom’s capabilities for real-time inference tasks. The initiative aims to reduce reliance on third-party providers like Nvidia, a critical step in controlling costs.
Additionally, OpenAI secured a $1.5 billion tender offer from SoftBank, providing liquidity for employees and reinforcing investor confidence. Speaking during SoftBank’s FY24 Q2 Global Conference Call, CFO Navneet Govil justified the valuation, stating, “It is the fastest consumer app to reach 100 million users…scaling from $28 million to $3.7 billion in revenues within two years.”
Competition Intensifies in the Generative AI Sector
As competition in generative AI continues to heat up, OpenAI is facing major pressure from rivals such as Google and Anthropic. Google’s Gemini model, which features advanced multimodal capabilities, and Anthropic’s Claude models are positioned as direct competitors to OpenAI’s offerings. And Elon Musk also entered the arena with his still young AI venture xAI, which just announced plans to expand its Colossus supercomputer tenfold, scaling AI capabilities with over 1 million GPUs
In response, OpenAI is expanding its product ecosystem. The company is reportedly developing the Orion model, a multi-step AI agent designed to enhance automation and reasoning capabilities, and exploring the creation of an AI-focused browser to challenge Google Chrome. These moves shows OpenAI’s ambition to maintaining its leadership position in a market projected to exceed $1 trillion in annual revenue within the next decade.
Ethical Dilemmas and Leadership Changes
OpenAI’s potential revision of its AGI clause has sparked debate over the ethical implications of its evolving business model. Originally founded as a nonprofit, OpenAI transitioned to a capped-profit model in 2019 to attract investment while maintaining a focus on ethical AI development.
However, critics, including Elon Musk, argue that this shift prioritizes financial gain over the company’s founding principles. Musk’s ongoing lawsuit against OpenAI challenges its governance structure, alleging undue influence by Microsoft.
Related: OpenAI: Released Emails of Elon Musk and Sam Altman Reveal What They Actually Discussed
Internally, OpenAI has experienced notable leadership changes in 2024, including the departures of CTO Mira Murati and Chief Research Officer Bob McGrew. These transitions have coincided with a period of rapid growth and strategic transformation. However, the return of co-founder Greg Brockman after a three-month hiatus has been viewed as a stabilizing force during this critical phase.
