Proton's AI manifesto for software engineering

Proton’s engineering manifesto argues that AI has permanently changed software development: models can now handle much of the code writing, testing, documentation, and boilerplate that once consumed engineers’ time. The shift is presented as an urgent business necessity, not a bid to follow fashion; companies investing in AI are shaping the competitive landscape, and Proton risks undermining its mission if it fails to adapt. The change does not make engineering expertise or sound practices obsolete. Engineers still bring judgment, accountability, and the ability to catch errors or malicious inputs, while maintainable architecture, documentation, and tests remain important. As these activities become cheaper to produce, however, established tradeoffs - such as how much testing is worthwhile - may need to change.

The manifesto identifies three obstacles to realizing AI’s benefits. First, engineers must move from writing code themselves to directing AI agents, making their work resemble team leadership: they should enable models to operate more independently, improve shared prompts and repository guidance, and address recurring failures. Second, teams must advance beyond using AI to speed up existing tasks and rethink the entire development process, finding and removing new bottlenecks in areas such as CI, design, approvals, and review. Third, this broader transformation requires coordinated company investment; individual productivity gains are not enough when teams lack the tools, resources, and organization to redesign shared systems. The central call is for engineers and Proton to take an active role in this transition rather than treat AI as a mere coding aid.