Software Industry Warns 1-Month Agentic Coding Boom Could Trigger Premature Governance

2 articles · Updated · InfoWorld · Jun 10

Within months of serious AI code use, developers have moved from quietly using tools like Claude Code to widespread “tokenmaxxing,” raising fears that governance will be imposed before practices mature.
The concern is not only speed but control: teams often lack clear oversight of which tools are used, how much is being spent, and whether the output delivers enough value to justify the cost.
Developers argue software deployment governance worked because it evolved from hard-won operational lessons, later formalized into documented processes rather than dictated upfront.
Top-down rules risk lagging a fast-moving technology, prompting calls for practitioners closest to the tools to shape standards before outside mandates lock in premature controls.

Sources

Will developers self-govern AI's risks before regulators are forced to step in?

Is a security catastrophe inevitable with AI generating dangerously flawed code?

As AI coding costs soar, are companies just buying digital snake oil?