SoftBank, the Japanese technology conglomerate, has rattled markets by completely selling its multibillion-dollar stake in Nvidia.
The transaction, reportedly netting SoftBank approximately $5.8 billion, was a clear signal that the company, under Masayoshi Son, is shifting its massive attention and capital toward an even bigger, perhaps riskier, bet on the future of artificial intelligence. This dramatic move, selling out of the world’s most valuable company, immediately sparked fears and debate over whether the current soaring AI market is heading for a perilous bubble.
Also read: SoftBank to invest $10-20 Mn in Perplexity AI at $3 Bn valuation
SoftBank’s rationale for shedding its Nvidia shares, considered the undisputed leader in AI hardware via its powerful GPUs, appears to be a strategic pivot. While Nvidia has dominated the infrastructure layer, providing the silicon backbone for the AI revolution, SoftBank is looking to invest aggressively in the application and model layer. Reports suggest the decision was heavily influenced by the emergence of new powerhouses like OpenAI, the creator of ChatGPT.
For SoftBank’s Vision Fund, which seeks to identify and fund transformative technology leaders, holding Nvidia was a successful investment, but selling it unlocks capital to chase the next paradigm shift. The move embodies a classic Son manoeuvre: cashing out on a realised win to double down on a perceived future mega-trend, shifting focus from the ‘picks and shovels’ (Nvidia’s hardware) to the ‘gold rush’ itself (foundational models and AI services).
The implication for the broader AI landscape is profound. SoftBank’s action suggests that the primary value creation in AI may be moving beyond hardware provision. While Nvidia’s chips remain essential, the market may be maturing to a point where the greatest returns lie in the companies building next-generation large language models, specialised AI chips, or advanced robotics, all sectors where SoftBank has historically sought influence.
This strategic divestment challenges the market’s current fixation on GPU supply, redirecting focus toward who controls the data, the models, and the proprietary applications that leverage this hardware. It signals that even the most astute investors believe the AI value chain is evolving rapidly, and yesterday’s champions may not necessarily be tomorrow’s ultimate beneficiaries.
Naturally, the colossal sale has intensified the simmering debate about an AI bubble. When one of the industry’s most prominent and historically aggressive investors sells a massive stake in the market’s leading stock, it inevitably raises questions. The current AI ecosystem is characterised by sky-high valuations, often based on potential future earnings rather than current profits. Massive capital injections into fledgling AI startups, combined with the rapid, often breathless pace of technological advancements, echo the euphoria of the late 1990s dot-com era.
However, many analysts believe the current situation is distinct. While a severe correction for certain overheated stocks and private valuations is plausible, the core technology of artificial intelligence is fundamentally real and rapidly penetrating every sector of the global economy.
Unlike the dot-com bubble, which saw many companies built on speculative business models, the demand driving AI—namely efficiency, automation, and discovery—is concrete. SoftBank’s decision is not an exit from AI but a reallocation of capital within it, moving from a stable infrastructure investment into high-growth, high-risk, unproven ventures.
This signals market volatility and potential corrections, but not necessarily a complete and total collapse of the underlying technological revolution. The gamble is whether Masayoshi Son has sold out at the peak of the hardware cycle to catch the beginning of the application cycle.
