Wedbush: DeepSeek Is a Positive Catalyst, Not a Threat, for These AI Software Stocks

Wedbush's note is a direct pushback against the panic that gripped AI stocks after DeepSeek's release. Rather than treating cheaper models as a threat to the economics of US AI software companies, the analysts frame it as an accelerant: lower inference costs mean enterprises can afford to run more AI-powered workflows, which increases demand for the software layer. The note specifically positions Palantir, Microsoft, Google, Amazon, Oracle, Salesforce and Snowflake as the names that stand to gain.

The "DeepSeek moment" refers to the release of the Chinese AI model that triggered a sell-off in US AI hardware and software stocks on fears that massive compute spending was no longer necessary. Wedbush's argument is that this fear is misplaced for software companies, whose business models depend on selling applications and platforms rather than on the cost of training frontier models. In this view, cheaper models expand the addressable market.

Wedbush also points to 2026 as the year the AI software trade broadens beyond the initial winners. The firm identifies enterprise AI adoption announcements and quarterly earnings as the key catalysts that will determine whether this thesis plays out.