I'd argue OpenAI weaning off Microsoft is a sign of strength. They no longer need give up immense future profits to a big tech company to back them, they have shown the ability to raise again and again. They just had the largest funding round in history. They are also probably the fastest growing company revenue wise in human history. Doom and gloom seems a bit premature.
Ignoring the AGI question (which I don't think your post is implying), I think this depends on whether we're counting success as having the best model or having a successful business. The latter they seem to be only extending their lead so far, from what I can tell.
I thought they were in trouble last year, as Anthropic had the clearly-superior model for so long. Yet normal people didn't care at all, and still barely know the words "Claude" or "Gemini".
OpenAI are executing very well on the consumer product side of things^1, and from what I can tell that's the side that actually matters to non-enthusiasts. Non-enthusiasts don't seem to push the models enough to notice the difference between the SOTA ones, so a "slightly better" model isn't enough to switch^2.
OpenAI also seem to be taking the bet (judging from Sam's interview on Stratechery) that features such as memory will create lock-in of users. Users wont want to switch to another bot that doesn't know them and their history very well.
I agree that the biggest risk is integration with existing tools becoming good enough that people don't install a separate app — Microsoft will probably own the business market once they integrate well enough that workers stop manually copying private data into ChatGPT. Though they'll likely still do so for the more "personal questions" about work. Probably the biggest risk for OpenAI's consumer-focus is closed platforms like Apple's pushing their own more-convenient AI, if they ever do. Or Meta in WhatsApp-dominant countries. They could use the built-in knowledge of you via your messages, etc, going back many years, though that's sometimes seen as more invasive than a bot that you told the information to yourself.
^1. Consumer features they lead: basic app quality/speed and convenience, memory, voice mode, image generation, customisability, deep research (execution and readability compared to the latest gemini pro), working "everywhere".
^2. Also see chatbot arena, where scores don't seem great at distinguishing intelligence beyond homework problems, and llama was able to "game" by praising the user more. Yet I would expect arena judges to be more enthusiast than the typical consumer.
It wasn't long ago that OpenAI was the undisputed leader in the AI race. Despite this, they appear to be struggling to remain competitive. I believe they've hit a tipping point and must either take drastic action to avoid being overtaken by rapidly advancing competitors and the structural advantages of big tech or start planning their out.
There is no doubt that OpenAI had an edge when they released ChatGPT. It was a hit. Apparently it surprised even them, and it left their competition floundering. Google tried to get people to care about Bard, Meta started cramming Meta AI into everything they could, Microsoft jumped on the OpenAI bandwagon, and everyone else dropped everything to join the fray. Despite this, my assessment is that OpenAI are no longer frontrunners in their domain.
Frontrunners No Longer
OpenAI are being carried by the wave of success they built and riding it out. I found myself wondering, "If OpenAI launched today, with the models they currently have, would they be cared about as much as they are now?" There is no doubt in my mind that they would certainly be cared about, but I'm not so sure they'd have such the recognition or status they have now.
The walls around OpenAI's moat are crumbling. They're certainly doing impressive work, and they are most definitely still competitive for now, but the lead they once had has disappeared, and at their current trajectory they'll be falling behind quickly. Many of OpenAI's researchers have moved on to roles at other companies, such as Anthropic.
Publicly available information and the long gap since significant architectural updates suggest that progress on GPT-5 may have encountered significant challenges or shifted priorities. The lack of a major public leap in core model capabilities gives the impression of slower progress compared to the rapid advancements seen earlier, or potentially a focus on areas like multimodal integration or safety that take more time. OpenAI has managed some decent improvements, but nothing to push them significantly ahead, and the competition has caught up and begun encroaching.
They've found themselves in a Half-Life 3 situation. It has gained a mythos beyond reality, and Sam Altman only continues to build it up further by throwing hype at hyped. No matter what GPT-5 is like, it won't meet the legendary expectations that it has garnered. At least Valve hasn't continued to hype it.
I think OpenAI are very aware of their current predicament. In Sam Altman's recent letter to employees regarding OpenAI's structure, he wrote:
Microsoft and OpenAI have been quite openly estranged, and based on that, I think Microsoft sees the writing on the wall. OpenAI has been useful for getting Microsoft a foot in the game, but now Microsoft is seeing the competitive viability of alternate models and shopping around.
DeepSeek came as a shock to Western AI companies, and the recent release of Qwen 3 has only kept that fire raging. Meanwhile Meta has been pushing improvements with Llama. Open models are decently comparable with most of OpenAI's offerings, meaning there isn't really a reason to choose OpenAI other than knowing of ChatGPT or having an existing relationship.
The only way OpenAI could have won the David vs Goliath battle long-term was if they built a moat too large for the competition, making it impossible for them to catch up. The issue is, they haven't just caught up; they've surpassed them.
OpenAI has been rushing and taking rash actions, as we've seen with the recent 4o sycophancy fiasco. It would not surprise me if they took drastic action in the very near future to convert.
OpenAI doesn't really have much sense of a lock-in. In fact, it is common for people who actively use AI and that care about the technology to use multiple models and switch between providers based on their needs. People jump around, and there isn't much stopping them from jumping away from OpenAI. Most APIs even come with a compatibility layer for OpenAI's.
The New Guard
Outside of this AI bubble (and let's not disillusion ourselves; we are very much in a bubble having this level of knowledge and experience), the majority of people aren't intentionally going out of their way to use chatbots. They are using AI where it is implemented as a convenient feature.
Google has what is widely regarded as the most capable set of models with Gemini 2.5 and a huge following already. Google Search is still extremely popular despite its waning quality, Google Drive is extremely popular, Gmail is the most popular mail provider, and Google Docs are used extensively. Not to mention Google's other products and their ownership of Android.
They are extremely well positioned with not only some of the most capable models, but some of the most applicable products. They're also Google, which means they've got insane economies of scale and fantastic existing business partnerships with a lot of companies.
Microsoft is in a similar position, although currently without a frontier model due to their alliance with OpenAI. Though I wouldn't be surprised if they managed to catch up decently quick given all the MAI rumours.
Big tech don't need to be perpetually ahead in raw model performance if they have comparable or slightly inferior models they can deeply integrate into the platforms they already have billions of people using daily.
I'm uncertain if OpenAI will go ballistic in an effort to remain afloat or simply become one of many AI companies while others step up and take their place at the adult table. Of course, every point I've made here could very well be out the window if a breakthrough occurs, but as it stands upon publication, I'm not sure how stable OpenAI's footing is.