‘Marking their own homework’: Sunak calls for independent AI testing amid job concerns

Artificial Intelligence is severely affecting the job market for young people, a fact of which we are all quite aware. This shift is mounting pressure on individuals regarding what the future holds for them.

To address this, Sunak, an advisor to Anthropic and Microsoft, shed light on current turmoil and was quite intrigued regarding the transformative impact of AI. 

He noted that company bosses have been privately consulting him about the recruitment of young people in the wake of this technology.

“There are reasons to be worried and think about the future. But we are able to do something about this,” he said.

“They're talking about this concept that they think they can continue to grow their businesses without having to significantly increase employment because they're starting to see how they can deploy AI," said Sunak.

“That's why I think we do have to look at this issue very seriously and with purpose."

The former chancellor told BBC Newsnight: “We should be thinking about, well, how do we tip the balance in favour of AI being used in that positive way… to help people do their jobs better [rather than replacing them].”

Sunak believes that most countries still need to analyse how to maintain balance in their systems as they face decreasing revenue from employment taxes and have to find that money from somewhere else. 

The significant impact on employment by AI means that it “may be different from previous technology cycles, and we want to do what we can to tip the scales in a more positive direction.”

Anthropic announced its new AI model, called Claude Mythos earlier this month. AI powerhouse Anthropic has officially entered the professional design space with the launch of Claude Design, a groundbreaking generative design workbench.

Powered by the newly released Claude Opus 4.7 model, the tool marks a significant strategic pivot for Anthropic as it moves from being a model provider to a direct competitor of software giants like Figma, Canva and Adobe.

Claude Design allows users to build and edit multi-page pitch decks, interactive prototypes, and website drafts through natural language.

Users can refine designs using chat, inline annotations or AI-generated sliders to adjust layout and style.

The tool can ingest a company’s existing codebase and design files to automatically extract and apply brand-consistent design systems including fonts, colors and components.

The company has clarified that it found the tool that outperforms humans in wake of ongoing hacking and cybersecurity threats. 

These tasks prompted by regulators and financial institutions have highlighted the risks to digital services. 

In this regard, Sunak said that his concerns regarding the development of Mythos showed that “we shouldn't rely on companies to mark their own homework.”

He further noted that it was to the UK’s credit that the AI Safety Institute under his leadership was the one to test Mythos’ capabilities. From his perspective, he has a firm belief in “Londomaxxing” and “Brit-maxxing” terms described within the tech industry as a surging wave of multi-billion pound investments in the sector.

“I know people like to talk us down or say that things aren't working but in this area is how some in the tech industry are describing a wave of recent multi-billion pound investments in the sector.”

Nonetheless, he believes that the UK has the opportunity to be the world’s most productive user of AI, leveraging the presence of prominent companies such as DeepMind, Anthropic and OpenAI in Britain.