The Hacker News thread debates a New York Times report that Meta may be claiming research-and-development tax credits for AI data-center hardware used largely in production rather than qualifying experiments. The classifications remain untested, potentially subject to IRS review, and have saved Meta billions, particularly about $3.9 billion in 2025. The top comment argues that legal tax avoidance reflects incentives created by federal policy, so blame belongs primarily with lawmakers rather than companies. Most replies reject that framing as too forgiving. Participants argue that corporations actively lobby for favorable rules, legality does not imply morality, and wealthy actors can be blamed for exploiting self-serving policies. Others contend that AI infrastructure may genuinely be experimental, that Meta pays substantial taxes overall, and that tax credits can encourage innovation. Broader disagreements cover corporate moral agency, media incentives, enforcement, social norms, lobbying, and whether tax reform should target corporate income, capital gains, or transaction taxes.