In his first speech since AI was elevated to the status of a fully-fledged ministry, Kanishka Narayan has warned that London will no longer rely solely on voluntary commitments from businesses.
On Monday 12 October, in west London, Kanishka Narayan gave his first speech since artificial intelligence became a ministerial-level portfolio in the British cabinet. His message to developers of cutting-edge models: control their autonomous systems more strictly, strengthen their safeguards and be more accountable for security. The government, he suggested, is ready to legislate if companies don't follow suit.
Towards mandatory reporting
Among the avenues being studied by ministers are the obligation to report incidents, particularly when an AI agent behaves maliciously, as well as tests imposed before being placed on the market. No measures have yet been decided. The speech comes a few days after a briefing by the heads of British intelligence (GCHQ, MI5, MI6) to the cabinet on the threats that AI poses to national security, and is accompanied by a section on the hardening of the country's cyber defences.
The international context also weighs heavily on the country: several countries, including Finland, France and Germany, have signed a call for common safety standards, which neither the United Kingdom nor the United States have joined. Moving to mandatory reporting would mark a step towards supervision enshrined in law, and a departure from Washington's preference for self-regulation.
Read the full article on the original source
Go to the source →