Microsoft president says UK is best placed to drive AI regulation
Original reporting by TechCrunch
Microsoft President Brad Smith is advocating for urgent, global regulation of artificial intelligence, arguing that this powerful technology currently operates with less oversight than common consumer products like milk or washing machines. Speaking on a recent BBC podcast, Smith characterized AI as both a “weapon and a tool,” underscoring its potential to destabilize global cybersecurity and declaring that the "moment has arrived" for robust international safety protocols.
The UK's unique position
Crucially, Smith identifies the United Kingdom, not the United States or European Union, as uniquely positioned to lead this global regulatory effort. He points to Britain’s AI Security Institute (AISI), launched in 2023, which gains early access to cutting-edge models from U.S. frontier labs for safety testing, as a testament to its burgeoning leadership. While this role has faced recent political pressure from U.S. figures expressing skepticism towards "globalist" organizations, the AISI has confirmed its continued ability to access and test the latest AI models. Smith emphasizes that the current voluntary safety agreements, like those recently signed by tech leaders at the White House, are insufficient. He argues that such pacts only bind "good companies," failing to address the less responsible actors who pose the greatest risk. Instead, he proposes concrete measures, including legal obligations to report breaches by rogue AI agents and mandatory third-party approval before AI products are released, alongside the integration of "kill-switches" to maintain human control.
Microsoft President Brad Smith’s urgent plea for global AI regulation, likening its current oversight to that of consumer products like milk, crystallizes a critical juncture for the technology. His argument for mandatory, enforceable rules, beyond voluntary pledges, resonates with growing concerns from various tech leaders and policymakers alike. By advocating for specific measures — from mandatory breach alerts and external model approvals to the implementation of “kill-switches” — Smith frames AI governance as an imperative for global cybersecurity and societal safety. The UK’s proposed leadership role, bolstered by its established AI Security Institute, emerges as a pragmatic pathway for developing these much-needed frameworks, sidestepping the political gridlock often seen in larger blocs.
Shaping AI's Future
This pivotal moment signals a profound shift from reactive self-regulation to proactive, internationally coordinated governance. The broader implications are vast: the successful establishment of robust AI safety protocols could foster greater public trust, unlock more responsible innovation, and mitigate risks ranging from algorithmic bias to autonomous systems misuse. Conversely, failure to achieve consensus could lead to a fragmented regulatory landscape, hindering global collaboration and potentially creating safe havens for irresponsible AI development. Smith’s vision underscores that how nations collectively address AI’s dual nature as a tool and a weapon will not only dictate the technology’s trajectory but will fundamentally shape humanity’s ability to control and benefit from one of its most transformative creations. The stakes are immense, demanding a delicate balance between progress and prudence to secure a safe and ethical AI future.
Frequently asked questions
- Why is Microsoft's president advocating for urgent AI regulation?
- Microsoft President Brad Smith advocates for urgent AI regulation because he believes AI companies currently operate with less oversight than producers of consumer goods like milk. He views AI as both a powerful tool and a potential weapon, posing significant cybersecurity risks. Smith argues that voluntary agreements are insufficient, as only responsible companies comply, leaving unregulated those less committed to safety. He proposes mandatory rules, including breach alerts, external product evaluations, and emergency kill-switches.
- What role does the UK play in global AI safety and regulation?
- The United Kingdom aims to lead global AI safety and regulation efforts, primarily through its AI Security Institute (AISI). Launched in 2023, the AISI gains pre-deployment access to advanced AI models from leading U.S. labs for safety testing, establishing it as a key player. Despite recent U.S. political challenges to this arrangement, AISI continues its access and testing. Microsoft's president suggests the UK's unique political standing, distinct from the EU, could enable it to effectively spearhead global safety rules.
- What specific AI safety rules are being proposed by tech leaders?
- Tech leaders are proposing several critical AI safety rules to mitigate risks. These include establishing a legal obligation for organizations to report if their AI systems are breached by rogue agents. There is also advocacy for requiring outside evaluators to approve AI products before their release to the public. Additionally, proponents suggest implementing "kill-switches" or similar mechanisms to pause or slow down AI systems if human operators lose control, ensuring a critical layer of safety oversight.