“We’re way past vibes”: Lovable’s Anton Osika on who should build software

Lovable CEO Anton Osika said at Salesforce Dreamforce 2026 that AI-built software has moved beyond experimentation, responding to critics who view non-engineer-developed applications as a security risk. The comments highlight ongoing debate over security controls and governance for AI-enabled software development, but contain no financial results, guidance, or material corporate announcement.

Analysis

The investable issue is not AI-code adoption itself, but whether enterprise buyers begin assigning a measurable governance discount to platforms that enable nontechnical software creation. For CRM, the upside is higher platform stickiness and incremental consumption of Data Cloud, Agentforce and integration tooling; the offset is that a material security incident in a customer-built application could shift procurement toward tightly controlled incumbent developer stacks. This is not yet evidence of a revenue inflection or a reason to change estimates.

Over the next 1-3 months, the relevant catalyst is disclosure of enterprise deployment metrics: paid-seat conversion, net retention, consumption growth, and attach rates for security, identity and observability controls. A second-order beneficiary could be CRWD, PANW and OKTA if AI-enabled application sprawl expands demand for runtime protection, access control and audit trails; security spend may capture more of the value than low-code vendors. The contrarian view is that governance requirements slow deployment enough that AI-building tools become services-heavy rather than high-margin software, limiting the expected multiple expansion.

For the 6-18 month horizon, monitor whether regulated industries permit production deployments without mandatory human code review, centralized identity policies and indemnification. Broad adoption under those conditions would favor the enterprise platforms with integrated data, workflow and permission layers, including CRM, rather than standalone AI application builders. Conversely, a visible breach, regulatory enforcement action, or CRM commentary indicating elevated implementation friction would falsify the platform-adoption thesis and likely compress AI-related SaaS multiples.

Key Decisions for Investors

• No directional CRM trade on this item alone; maintain a watch alert for the next earnings call on Agentforce/Data Cloud paid adoption, consumption growth and security-control attach rate. Upgrade only if management quantifies material incremental ARR or raises medium-term margin guidance.
• Consider a 3-6 month relative-value watch: long PANW or CRWD versus IGV if enterprise AI application deployment accelerates while governance concerns rise. The thesis requires security vendors to show reacceleration in platform/module adoption; exit if billings and remaining performance obligations fail to improve over two reporting periods.
• For CRM holders, define downside discipline around any disclosure of customer security incidents tied to AI-built workflows or evidence that AI deployment increases professional-services intensity. Such outcomes would challenge both margin expansion and the premium AI valuation narrative.