Feryal Clark built her political standing in Hackney rather than Westminster. During six years overseeing transport and the environment, she backed road closures, cycling infrastructure, School Streets and restrictions on heavily polluting vehicles. The policies were contentious, but they gave her something many aspiring MPs lack: a record of making decisions and defending them.

She secured the Enfield North nomination in 2019 after the sitting MP, Joan Ryan, left Labour. The selection was not entirely clean. Activists protested that Labour's national executive had restricted the choice of candidates, and 46 of the 227 ballots returned were spoiled. Clark still won convincingly among the valid votes.

Her internal politics placed her on Labour's pragmatic soft left. She supported Lisa Nandy rather than Keir Starmer or Rebecca Long-Bailey in the 2020 leadership contest. Starmer nevertheless promoted her through shadow health and crime briefs before appointing her minister for AI and digital government after the 2024 election.

What she stands for

Clark's most consistent politics combine public-service reform with environmental intervention. In Hackney, she was prepared to restrict car use to improve air quality and encourage walking and cycling. By 2017, more than 100 borough roads had been closed to through traffic, while the council later introduced School Streets and the country's first ultra-low-emission street zones. These were collective council policies, but Clark controlled the transport portfolio when they were developed and implemented.

She did not avoid confrontation. When residents argued that one road-closure scheme would divert pollution past a primary school, Clark accused campaigners of preferring "publicity stunts" to evidence. She later admitted that the council's initial traffic assumptions had been underestimated and that air-quality modelling was late. The policy objective was defensible. Her handling of residents was less impressive.

As AI minister, she presented technology as a way to reduce administration, improve medical diagnosis and make public services more equal. Her approach favoured adoption, voluntary standards and cooperation with industry over immediate statutory regulation.

What she actually delivered

Clark's clearest completed national policy was the AI Cyber Security Code of Practice, published in January 2025. It set baseline security principles for organisations developing or operating AI systems and was intended to form the basis of an international standard. It was guidance, not a legal obligation.

Her department also produced a voluntary Software Security Code of Practice and began testing "Humphrey" AI tools with 25 councils. Early research suggested that an automated minute-taking tool could save roughly an hour of administrative work after an hour-long meeting. That was a pilot finding, not evidence of savings across local government.

The first public version of the GOV.UK app launched while she held the digital-government brief. It offered personalised links and search functions, with plans for notifications, an AI assistant and integration with a digital wallet. At launch, it remained a basic gateway to existing webpages. Even Technology Secretary Peter Kyle called it a "minimum viable product".

These were real steps, but none amounted to the transformation promised in government publicity. The codes were voluntary, the council programme was experimental and the app's most useful functions had not arrived.

What she did not deliver

Labour entered office promising legislation requiring developers of powerful AI systems to submit their models for safety testing. The bill was repeatedly delayed while ministers reconsidered their approach and sought closer alignment with the United States. It did not become law during Clark's ministerial tenure.

She also left the government without resolving the fight between AI companies and copyright holders. The department's preferred proposal would have allowed AI developers to train models on protected work unless creators opted out. Artists, publishers and parliamentary committees argued that this placed the burden on the people whose work was being used. Clark said the government would not proceed until it had an effective system, but no settlement emerged before she departed.

Clark left ministerial office in September 2025 after 14 months. No official explanation was given. She described the change as stepping back, while the government recorded only the end of her appointment. Her appointment as trade envoy to Germany in January 2026 showed that she retained the leadership's confidence, but the unpaid role had produced no published trade or investment result by the end of July.

Capability judgement

Clark's strongest evidence of capability comes from local government. She pursued identifiable transport and air-quality policies, implemented them and accepted political opposition, although her treatment of critics could be abrasive and some modelling proved unreliable.

Her national record is thinner. She oversaw useful cyber-security standards, public-service trials and the first version of the GOV.UK app. The difficult work remained unfinished: enforceable AI safeguards, a workable copyright settlement and proof that the technology produced the savings ministers claimed.

She has demonstrated practical ability as a local policymaker and competence as a junior minister. She has not yet demonstrated that she can carry a major national reform from promise to completed result.