US and UK sign agreement on AI safety

In a landmark agreement, the United States and the United Kingdom have pledged to collaborate on testing advanced AI safetyIn a landmark agreement, the United States and the United Kingdom have pledged to collaborate on testing advanced artificial intelligence (AI). This is the first-ever bilateral deal of its kind focused on AI safety. This agreement builds on commitments made at the AI Safety Summit last year, where both countries established AI Safety Institutes. These institutes will work together to develop robust methods for evaluating the safety of AI tools and systems.

“This is the defining technology challenge of our generation,” said Michelle Donelan, the UK’s tech minister. She emphasized the importance of international cooperation, stating, “Only by working together can we address the technology’s risks head-on and harness its potential to improve our lives.”

While the AI development field has been booming, concerns about safety linger. The biggest AI chatbots, like ChatGPT, Gemini, and Claude, are locked in fierce competition. While the companies behind them seem open to regulation, governments haven’t yet implemented significant controls. Additionally, access to crucial information, such as training data and environmental impact, remains limited.

The EU’s upcoming AI Act, however, will require developers to be more transparent about risks and data usage. This follows OpenAI’s recent decision to shelve a voice cloning tool due to potential misuse, especially during elections. A similar incident in January saw a fake AI-generated robocall impersonate US President Joe Biden.

Currently, most AI systems perform specific tasks, but fears exist about the potential dangers of more advanced “general” AI. Professor Sir Nigel Shadbolt cautions that AI, like other scientific advancements, can be used for good or bad. He believes focusing on research and understanding the potential risks is crucial.

“This alliance underscores the commitment of both nations to pioneer the safe evolution of artificial intelligence and marks a significant step towards a unified front navigating the complexities of AI,” said Flavia Colombo, VP EMEA at HubSpot. “It’s testament to the fact that in AI development, collaboration and safety are not just optional; they are imperative for ensuring that we continue to unlock AI’s transformative power across all sectors, from healthcare to cybersecurity, with the utmost integrity and foresight.”

US Commerce Secretary Gina Raimondo sees the agreement as a way to gain a deeper understanding of AI systems, allowing for better guidance. “This partnership shows we’re not shying away from these concerns – we’re actively addressing them,” she said.