SoftBank’s Masayoshi Son voices concern over AI safety risks

Masayoshi Son has spent years as artificial intelligence’s most enthusiastic cheerleader. Now the SoftBank Group founder is voicing concern about AI safety risks as machines become more capable, according to Bloomberg.

The optimist who wrote the biggest checks

Few investors have tied their fortunes to AI as tightly as Son. SoftBank’s total commitments to OpenAI are projected to reach $65 billion by October 2026.

Son has also refused to entertain talk of a bubble. At SoftBank’s annual general meeting on June 24, 2026, he said AI is “still in its early stages” and pushed back hard on investors who worried valuations had run too far, framing those bubble concerns as an insult to the technology itself.

He went further on timelines. Son has pulled forward his target for artificial superintelligence, which he defines as AI that is 10,000 times smarter than humans, to within two years.

Why safety talk moves SoftBank’s stock

The market has already shown how sensitive SoftBank shares are to this exact conversation. On September 14, 2026, the stock plunged more than 13% after leaders from OpenAI and Anthropic expressed caution about safety risks in frontier model development.

That was SoftBank’s steepest single drop in almost three months. Executives at OpenAI and Anthropic have also called for oversight of self-improving AI systems.

For SoftBank, that question carries extra weight because of how it has funded its ambitions. The company raised $11.1 billion through a junk bond sale this year to finance its AI push, at a time of rising borrowing costs.

A balancing act with a lot of leverage

If oversight of frontier AI tightens, development timelines could stretch. Costs could rise. The path to the returns SoftBank is counting on could become longer and bumpier, and leverage tends to amplify bumps rather than absorb them.

What this means

For SoftBank shareholders, the key takeaway is that AI safety is no longer an abstract debate happening in research papers. It is a live market variable that has already moved the stock sharply once this year.

Investors will likely watch three things closely. First, how AI labs like OpenAI and Anthropic continue to talk about frontier risks, since their statements have shown they can move SoftBank’s share price. Second, whether calls for oversight of self-improving systems turn into concrete regulatory action. Third, how SoftBank manages its debt load as borrowing costs rise and its AI commitments keep growing.

With $65 billion projected for OpenAI and $11.1 billion in fresh junk debt on the books, the margin for surprise is thinner than it once was.