Bill Gates is calling for changes to the way governments tax businesses using artificial intelligence, while Microsoft president Brad Smith has warned that safeguards are struggling to keep pace with the technology.
Their interventions come amid debate over how to protect workers from automation, prevent dangerous AI misuse and ensure companies remain accountable for the systems they develop.
Companies are under pressure to adopt automation, while policymakers must consider what happens to workers whose roles become less necessary. As AI systems become capable of carrying out more tasks autonomously, questions remain about how their actions should be monitored and who should be held responsible when things go wrong.
Gates is focusing on the economic consequences of automation, while Smith's proposals centre on AI safety and accountability. Their approaches raise questions about the role governments should play in managing the technology's effects.
Bill Gates Proposes AI Taxes as Automation Threatens Jobs
In an essay published on 26 August, Gates argued that governments should reconsider how they tax human labour compared with automated work. His proposal includes taxing AI tokens and robots, with the aim of reducing the financial incentive for employers to replace people with machines.
'The tax system nudges you toward replacing people with machines,' Gates wrote in his essay, published on Gates Notes. He argued that companies pay payroll taxes when they employ people, while robots can generally be written off immediately as business expenses.
The Microsoft co-founder says revenue from the proposed measures could help fund worker retraining and strengthen the social safety net for people affected by automation. He has also proposed reserving some jobs or tasks for humans, either to protect workers from disruption or to preserve the human element in sensitive situations.
Gates calls this idea 'Human Reserved'. Rather than banning AI from entire industries, he suggests that some roles could remain human-led, while others could adopt the technology gradually over several years or decades.
Education and mental health care are among the areas where he envisages people working alongside AI rather than handing over responsibility entirely.
However, Gates acknowledges that these proposals would be difficult to implement. Governments would need to decide which occupations qualify, how to prevent businesses from bypassing restrictions and how to handle competition from countries that allow greater automation.
He also recognises that taxation must be designed carefully. A measure that discourages companies from replacing workers should not unnecessarily obstruct useful applications that could make healthcare and education more affordable.
Most recently, Gates addressed the risks of AI misuse in a post on X on 11 October 2026, writing:
'Just as we regulate planes, cars, and medicines, we need safeguards that help us realize AI's incredible potential while reducing the risks. They won't eliminate every harm, but they can help prevent misuse by bad actors. We're not where we need to be yet—and this deserves much more of our attention.'
Just as we regulate planes, cars, and medicines, we need safeguards that help us realize AI's incredible potential while reducing the risks. They won’t eliminate every harm, but they can help prevent misuse by bad actors. We’re not where we need to be yet—and this deserves much… https://t.co/SubCPPgSCp
— Bill Gates (@BillGates) October 11, 2026
Brad Smith Warns AI Safety Rules Need to Catch Up
In an interview with the BBC's Big Boss Interview, the Microsoft president called for stronger safeguards around powerful AI systems, including emergency mechanisms and legal reporting requirements.
'Right now, your washing machine is subject to more safety rules than AI. Every carton of milk that you see in the grocery store is subject to safety rules and inspection,' Smith said.
He described AI as both a 'tool and a weapon' and called for immediate action. Among his proposals are emergency mechanisms that can stop an AI system if it behaves dangerously, alongside legal requirements for companies to report serious incidents and inform affected parties when an AI agent breaches systems.
Smith also raised the possibility of independent government evaluators checking particularly advanced AI models before their release. He argued that established safety practices used in other industries could help inform the rules applied to AI.
His proposals would place greater emphasis on enforceable requirements rather than relying solely on voluntary industry commitments. At a White House meeting involving major technology executives, participating companies agreed to work with independent auditors to assess whether their systems operate as intended and avoid unauthorised access to technical systems.
Smith welcomed that commitment but said it did not go far enough. 'It's a good step, it's a necessary step... But I don't think it's sufficient any longer,' he said.
He also warned that voluntary oversight has an inherent weakness: responsible companies may follow the rules, but less responsible developers could choose not to.
Smith believes the UK could help coordinate international standards, citing the technical capabilities of its AI Security Institute. He also suggested that Washington might be more receptive to British-led efforts than rules originating in Brussels.
