Show the Working

Create New Online Crime Agency

Liberal Democrat · what the evidence says

An independent, source-checked look at Liberal Democrat’s policy “Create New Online Crime Agency” — what it would actually do across the things that affect your life. Every claim below quotes the source behind it. How this works.

Personal liberty & free speech — Little effect

minor · low confidence

The agency targets activity already illegal — fraud, revenge porn, threats — so it does not extend state coercion over lawful speech or choices. The one liberty-adjacent concern is expanded data-sharing between agencies, but the evidence gives no detail on new surveillance powers over ordinary citizens.

The evidence

Biggest unknown: Whether the agency is granted new data-collection or interception powers that extend beyond existing law — nothing in the evidence confirms or rules this out.

Our reading: O10 is concerned with state coercion over lawful speech, bodily autonomy, surveillance and mandates. This policy targets conduct already criminalised — fraud, revenge porn, incitement — so it does not extend prohibitions into new expressive territory. The stated aim is operational efficiency for local policing, not an expansion of the categories of illegal speech. The one plausible liberty-adjacent mechanism is the intelligence-sharing model between agencies, banks and tech firms (E4, E16), which could in principle involve broader data collection on individuals. However, the evidence provides no detail on whether new surveillance or interception powers are granted; it only notes that data-sharing barriers remain an open question. Without evidence that lawful speech, privacy or freedom from state coercion is materially affected, the effect on O10 is at most minor — and even that rests on speculation about implementation detail not grounded in the provided evidence. The direction is therefore negligible on current evidence, with low confidence because the absence of detail about the agency's legal powers is itself informative only up to a point.

Crime, justice & national security — Helps

minor · low confidence

Creating a dedicated Online Crime Agency targets the UK's most prevalent crime — fraud — plus other online harms, and could free up local officers for community policing. But funding is modest relative to the scale of the problem, and key implementation hurdles around data-sharing and technology obsolescence remain unresolved.

The evidence

Biggest unknown: Whether the agency can overcome legal and operational barriers to data-sharing between agencies and industry, and keep pace with rapidly evolving fraud methods.

Our reading: The policy targets fraud — the UK's most prevalent crime — plus revenge porn and online threats, all of which bear directly on O5 indicators (crime rates, antisocial behaviour, victims). The mechanism is credible: a dedicated agency with a public-private collaborative model, focused on disrupting criminal infrastructure at source rather than relying on post-incident policing. The NCA's recent trajectory (rising investigations and convictions) suggests institutional capacity is building. The secondary benefit — freeing local officers for community policing — adds a modest further O5 gain. However, the magnitude must be tempered. The £250 million three-year envelope is small relative to the £14.4 billion annual fraud cost (roughly £1 of public resource per £58 of social harm). Data-sharing barriers remain unresolved, past voluntary industry action has been slow, and rapid technological change threatens to outpace implementation. These are not fringe concerns — they are flagged by independent analysts. Absent the policy, the status quo involves fragmented policing of online crime and local forces absorbing online fraud caseloads. The agency represents a genuine structural improvement on that baseline, but the evidenced effect is real-but-limited: it is unlikely to move fraud prevalence dramatically at population scale within a parliament given funding constraints and implementation risks. Direction: improves; magnitude: minor; confidence: low given genuine implementation uncertainty.