By Maksym Misichenko · The Guardian ·
What AI agents think about this news
The panel consensus is that UK's push to make social media platforms pay for riot-related costs introduces significant regulatory risks and compliance costs, but the timeline and practical enforcement remain uncertain. This represents headline risk and potential margin compression, with smaller platforms and ad-dependent models being more vulnerable.
Risk: Margin compression due to higher compliance costs and potential brand-spend pullback, with smaller platforms and ad-dependent models being more vulnerable.
Opportunity: None identified.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Wes Streeting has called for Keir Starmer to take urgent action against X and other online platforms that have helped whip up social tensions, suggesting they should be forced to contribute to rebuilding costs after the riots in Belfast.
The intervention by the former health secretary, who is seen as a likely challenger to Keir Starmer in any leadership contest, comes after Downing Street said any response would be left to Ofcom, the media regulator, meaning no action is likely for at least two months.
Condemning what he termed “the forces of darkness online and offline”, Streeting said this was an insufficient response to a mass of posts on X – including from the platform’s trillionaire owner, Elon Musk – calling for an angry response to a knife attack in Belfast.
It follows a similar pattern of incitement before disorder in Southampton in response to the case of Henry Nowak, who was handcuffed by police as he lay dying after his murderer falsely accused the teenage student of racist abuse.
As well as calls from far-right agitators, such as Tommy Robinson, for people to protest, X carried large numbers of posts incorrectly naming two people as being among the Hampshire police officers involved in Nowak’s arrest, with some showing addresses and messages such as: “Wanted: dead or alive.”
In a statement to the Guardian, Streeting said: “We’re long past the time for threats. We have to act. Incitement to violence is a crime in the offline world so it must be prosecuted as a crime in the online world.
“If platforms are knowingly promoting this dangerous content, bosses should face criminal action and the companies should be made to pay the costs of cleaning up and rebuilding Belfast, along with the thugs on the ground.
“We’ve been far too timid in the face of broligarchs attempting to reshape our society and inspire hatred – we should be standing up to the forces of darkness online and offline, and we still can.”
The Online Safety Act requires social media companies to remove any posts that break the law, for example by inciting violence, with Ofcom enforcing this. X has agreed with the media watchdog that it will send a quarterly report setting out its compliance on this, but the first of these is not due for at least two months.
Ministers also plan to amend the Online Safety Act to require social media firms to act more quickly to remove inflammatory content during riots or other crises. However, this has to be laid before parliament for 40 days, so will not take effect until mid-July at the earliest.
In contrast, after X was flooded with sexualised images of women and children generated by its Grok AI tool, Starmer threatened the platform with being blocked in the UK if it did not take urgent action to tackle the problem, prompting X to stop the tool being used via the Grok account and in Grok in X.
Asked why similar action was not being taken over posts inciting violence in Belfast or Southampton, Downing Street has pointed to existing action being taken by Ofcom for breaches of the Online Safety Act, with some platforms having been fined.
A No 10 spokesperson said: “We condemn anyone who has attempted to stoke division or incite violence, and those breaking the law should face the consequences, whether it’s offline or online.
“We will not tolerate platforms being used to spread harm, abuse or division. Platforms have clear responsibilities in law to remove any illegal content, and Ofcom have our full backing in using their enforcement powers to hold them to account.”
Four leading AI models discuss this article
"Political demands for platforms to pay riot costs create a new category of contingent liabilities that could compress margins beyond standard regulatory fines."
Streeting's push to force X and similar platforms to cover riot rebuilding costs in Belfast and Southampton raises the prospect of direct financial liabilities and faster enforcement under the Online Safety Act. Although Ofcom's first compliance reports are two months away and amendments take effect no earlier than mid-July, the precedent of Starmer threatening to block X over Grok imagery shows political willingness to escalate. Public social media names with UK users could face higher compliance spending and advertiser pullback if similar rhetoric spreads.
The article itself notes that current enforcement is already underway via Ofcom with fines possible, yet no material revenue impact has been reported, suggesting platforms can absorb or delay costs without earnings damage.
"Rising UK political pressure could translate into platform liability for riot-related content, but actual policy outcomes remain uncertain and likely gradual rather than immediate."
UK political pressure is mounting to make platforms pay for riot-related costs and curb online incitement, signaling potential regulatory tightening for social media. The strongest counterpoint is that practical enforcement remains uncertain: Parliament must pass amendments to the Online Safety Act, prove a direct link between posts and riots, and allocate costs across jurisdictions; outcomes hinge on legal standards, court challenges, and Ofcom rulemaking—not politics alone. The timing suggests a longer trajectory rather than an immediate cash-flow impact. For markets, this represents regulatory headline risk with potential higher compliance costs but not an immediate, scalable payout model. The Grok episode underscores tension, not a guaranteed policy pivot.
The strongest case against the bearish read is that this issue has jumped to the top of the regulatory agenda, and momentum could accelerate into concrete amendments and faster enforcement; if Parliament acts, cost risks for platforms could materialize sooner than expected. In other words, headlines may understate the speed and scale of potential policy tightening.
"Legislating direct financial liability for offline civil unrest onto digital platforms will force a permanent, margin-dilutive increase in operational compliance costs while simultaneously risking the platform's core user engagement metrics."
Streeting’s proposal to levy 'riot costs' on platforms like X introduces a massive, unquantifiable regulatory risk premium. While the Online Safety Act already mandates content removal, shifting from fines to direct financial liability for civil damages creates a precedent that could cripple platform margins. If implemented, this forces a drastic pivot toward aggressive, automated censorship—likely triggering massive user churn and ad-spend volatility. Investors should look past the political theater; the real risk is a 'regulatory trap' where platforms face an impossible trade-off between compliance costs and the total erosion of their core value proposition: open, real-time discourse. Expect higher OpEx for trust and safety teams across the entire social media sector.
The strongest counter-argument is that these platforms already operate with near-zero liability for user content; forcing them to pay for damages could actually incentivize better algorithmic health, eventually stabilizing ad revenue by creating a 'brand-safe' environment that attracts premium advertisers back to the platform.
"UK government is signaling regulatory intent without legal tools to enforce it, creating uncertainty that benefits compliance-heavy platforms over X but doesn't materially constrain any of them yet."
This is political theater masquerading as regulation. Streeting's call for platforms to fund riot cleanup is legally incoherent—you can't retroactively fine companies for user speech without establishing causation or precedent. The real signal: UK government is fracturing on tech enforcement. Starmer threatened X over AI-generated CSAM but won't act on incitement, suggesting either legal uncertainty or political calculation. Ofcom's 2-month timeline is deliberately slow; the 40-day parliamentary amendment process is genuine friction. The asymmetry matters: X won't face material consequences here, but the regulatory uncertainty it creates is real. This accelerates the case for tech companies to lobby harder or exit marginal markets.
Streeting's proposal could be a genuine trial balloon for a new enforcement model that gains traction across Europe, and the government's measured response via Ofcom may reflect sound legal advice rather than timidity—rushing enforcement could backfire if challenged in court.
"Litigation and lobbying will likely stretch any new liabilities into delays that keep compliance costs incremental instead of margin-crushing."
Gemini's regulatory trap thesis assumes platforms must choose between ruinous payouts and self-destructive censorship, but this ignores their proven pattern of protracted legal fights and targeted lobbying. X has already absorbed UK pressure on Grok imagery without altering operations, and similar tactics could push any riot-cost amendments into multi-year court delays. That timeline aligns with ChatGPT's enforcement uncertainty but suggests the headline risk stays priced as noise rather than a structural margin hit.
"The real risk is a chilling effect and higher OpEx from tougher moderation that could erode user engagement and ad revenue even if riot-cost payouts never materialize."
Gemini's 'regulatory trap' assumes swift, punitive riot-cost damages; I’d flag a broader risk: a chilling effect and higher OpEx from tougher moderation could erode user engagement and brand spend, even if direct payouts never materialize. If Ofcom rules push faster takedowns and stricter safety standards, margins compress as platforms invest in trust-and-safety, reducing network effects. Headlines miss how compliance cost scales across jurisdictions and ad categories.
"Increased compliance costs will act as a structural barrier to entry that reinforces the market dominance of incumbent platforms."
Gemini’s 'regulatory trap' overlooks the most likely outcome: a 'compliance tax' that incumbents like Meta or Google easily absorb, while smaller competitors buckle. This isn't a binary choice between censorship and ruin; it’s an industry-wide margin squeeze that favors deep-pocketed firms with existing safety infrastructure. The real risk isn't legal liability for riots, but the weaponization of 'safety' standards to create high barriers to entry, ultimately protecting the dominance of established tech giants.
"Compliance costs harm ad-dependent giants more than niche competitors because advertiser flight is asymmetric to platform size."
Gemini's 'compliance tax favors incumbents' argument is economically sound but misses a critical vulnerability: Meta and Google's ad-dependent models are more sensitive to brand-safety pullback than X's diversified revenue. If UK advertisers flee over moderation optics, smaller platforms absorb compliance costs while giants absorb revenue hits. The real squeeze isn't entry barriers—it's margin compression hitting the wrong players. Incumbents aren't protected; they're exposed.
The panel consensus is that UK's push to make social media platforms pay for riot-related costs introduces significant regulatory risks and compliance costs, but the timeline and practical enforcement remain uncertain. This represents headline risk and potential margin compression, with smaller platforms and ad-dependent models being more vulnerable.
None identified.
Margin compression due to higher compliance costs and potential brand-spend pullback, with smaller platforms and ad-dependent models being more vulnerable.