The UK government wants to create a tough pivot to boost the economy and industry with AI. As part of this, we pivot an institution that we founded with a very different purpose about a year ago. Today, the Bureau of Science and Industrial Technology announced that it will rename the AI Safety Research Institute to the AI Security Research Institute. (Same first letter: Same URL.) So, the body primarily explores areas of existential risks and biases in large-scale language models, and specifically “AI protects national security and risks against crime.” Focus on cybersecurity, focusing on strengthening. . ”
In addition to this, the government has also announced a new partnership with humanity. No solid services were announced, but the MOU shows that the two will “explorate” using humanity’s AI assistant Claude in public services. Humanity aims to contribute to the work of scientific research and economic modeling. The AI Security Institute also provides tools to assess AI capabilities in the context of identifying security risks.
“AI has the potential to transform how governments can serve citizens,” Dario Amodei, co-founder and CEO of humanity, said in a statement. “Human AI assistant Claude will strengthen public services with the aim of discovering new ways to enable UK residents to make critical information and services more efficient and accessible to UK residents. I look forward to exploring how I can help you with this.”
Humanity is the only company announced today, coinciding with a week of AI activities in Munich and Paris, but that’s not the only one that is working with the government. The set of new tools announced in January was all equipped with Openai. (At the time, Secretary of State for Technology, Peter Kyle, said the government was planning to work with a variety of basic AI companies, and that is what human trade proves.)
The AI Safety Institute’s government switch was launched over a year ago with many fanfares, but AI security is not too surprising.
It was noted that when the newly established Labour government announced a heavy change plan in January, the words “safety”, “harm”, “existence” and “threat” were not shown in the documents at all. It was worth it.
That was not an oversight. The government’s plan is to start investing in a more modernized economy to do so using technology, particularly AI. It wants to work more closely with Big Technology and also build its own homemade big technology.
The main messages it is promoting in support of it was development, AI, and more development. Civil servants are encouraged to have their own AI assistants called “Humphrey,” share data and use AI in other areas to speed up their work methods. Consumers get digital wallets for government documents and chatbots.
Have you solved the AI safety issue? It’s not accurate, but the message cannot be considered at the expense of progress.
The government insisted that despite the name change, the song remains the same.
“The changes I am announcing today represent the logical next step in how we approach responsible AI development. It will help unlock AI and grow the economy as part of our plan for change.” Kyle said in a statement. “The work of the AI Security Institute remains the same, but this new focus is that our citizens and our allies use AI for our institutions, democratic values and ways of life. It ensures you are protected from those who try to do so.”
“From the start, the lab’s focus has been on security, and we have built a team of scientists focused on assessing serious risks to our citizens,” said Ian Hogarth, who remains the lab’s chair. added. “Deepening our new criminal misuse team and partnership with the national security community identifies the next step in addressing these risks.”
Even further, priorities seem to have definitely changed regarding the importance of “AI safety.” The biggest risk that the American Institute of AI Safety is currently thinking is that it will be demolished. US Vice President JD Vance telegraphed earlier this week during his speech in Paris.
TechCrunch has a newsletter focused on AI! Sign up here to get it every Wednesday in your inbox.
Source link
