Meta and TikTok: What Whistleblowers Alleged About the Race for Engagement

Meta and TikTok faced renewed scrutiny in 2026 after more than a dozen whistleblowers and insiders told a BBC investigation that competition for attention had pushed major social-media companies to accept safety risks they knew existed. Reporting around the BBC documentary Inside the Rage Machine described allegations involving recommendation algorithms, Instagram Reels, moderation workloads and decisions about harmful or borderline content.

But the distinction between allegation and established fact matters. The whistleblowers’ accounts do not by themselves prove that Meta and TikTok deliberately set out to harm users or that every recommendation decision was made to maximise profit at the expense of safety. Meta and TikTok reject that characterisation and point to moderation systems, teen-safety settings and recommendation rules.

What has changed since the original March 2026 reporting is the regulatory context. By September, Ofcom had opened a formal investigation into TikTok’s compliance with UK child-safety duties, the European Commission had preliminarily found Meta in breach of the Digital Services Act over the addictive design of Facebook and Instagram, Pennsylvania had sued TikTok over alleged youth harms, and Meta had agreed to an $18 billion settlement with most US states over separate allegations involving children and social-media addiction.

That does not prove the whistleblowers were correct in every detail. It does show that the wider argument over engagement-driven design has moved from internal complaints into courtrooms and regulatory enforcement.

The BBC Investigation Put the Algorithm Race Under a Microscope

The March investigation, tied to the documentary Inside the Rage Machine, brought together accounts from more than a dozen insiders across technology companies.

At the centre of the Meta and TikTok story was a basic commercial problem: social-media platforms compete for time and attention.

Recommendation systems are designed to predict what a user is most likely to watch, like, share, comment on or continue consuming. That can make feeds more relevant and entertaining. It can also create incentives to prioritise emotionally intense material because outrage, fear, shock and conflict often generate strong reactions.

BBC journalist Marianna Spring later told a UK parliamentary committee that interviews conducted for the documentary indicated that engagement algorithms aim to capture attention, while negative emotions such as outrage, anger and upset can be particularly effective at doing that.

Whistleblowers told the BBC that internal research at Meta and TikTok had already exposed some of those trade-offs, although the evidence differed between the two companies.

A former Meta engineer, identified in reporting as “Tim”, alleged that senior managers told engineers to allow more “borderline” content into recommendation systems as Meta tried to compete more aggressively with TikTok.

Borderline content generally refers to material that may be disturbing, misleading or socially harmful without necessarily crossing a platform’s formal removal rules.

The engineer said the pressure was linked to engagement and competition, and he recalled being told that falling stock performance was part of the concern.

Meta disputes the suggestion that it deliberately amplifies harmful content for financial gain.

Why TikTok Changed the Competitive Landscape

The rise of TikTok changed what social-media users expected from a feed.

Instead of relying mainly on accounts a person had chosen to follow, TikTok’s For You feed became highly effective at recommending videos from people a user had never seen before. The system could learn interests quickly from viewing behaviour and then serve an almost continuous stream of personalised clips.

That model was enormously successful.

Competitors responded. Meta launched Instagram Reels in 2020 and pushed short-form video more heavily across its products.

Former Meta researcher Matt Motyl told the BBC that Reels was part of the company’s attempt to match the product experience TikTok had popularised. His work involved large-scale experiments on how ranking changes affected behaviour.

This is where the Meta and TikTok debate becomes less about one piece of content and more about the architecture of the platforms.

If a ranking system rewards posts that hold attention longest, and emotionally provocative material consistently performs well, a company has to decide whether to accept that engagement or deliberately reduce the reach of content that may be harmful without clearly breaking rules.

That is a product-policy choice as much as an engineering choice.

Internal Reels Research Raised Specific Safety Questions

One of the strongest pieces of evidence described in the March investigation involved internal research on Instagram Reels.

Documents shared by Motyl reportedly showed that comments on Reels had a higher prevalence of harmful material than comments elsewhere on Instagram.

The figures cited in reporting were striking: bullying and harassment were 75% more prevalent, hate speech 19% more prevalent, and violence or incitement 7% more prevalent in the material examined.

Those numbers need context. They came from internal company research shared by a former employee and described in the investigation. They should not be treated as a universal measurement of everything ever posted on Reels.

Still, they are important because they indicate that Meta’s own researchers were examining whether the rapid expansion of short-form video created different safety conditions from the rest of Instagram.

Another internal study described financial incentives created by ranking systems as potentially misaligned with Meta’s mission because creators could be rewarded for taking a path that maximised profits at the expense of audience wellbeing. The wording later appeared in Australian parliamentary material discussing the BBC investigation.

The central question raised by those documents was not whether algorithms can identify harmful material. It was whether business incentives encouraged the company to tolerate too much content near the edge of its rules.

Safety Staffing Became Part of the Whistleblower Case

The allegations against Meta and TikTok were not limited to algorithms.

Another former senior Meta employee told the BBC that hundreds of staff were added to help grow Reels while safety teams struggled to secure much smaller numbers of specialists for child protection and election integrity.

Reporting on the allegations said about 700 staff were added to the Reels expansion, while requests for two additional child-safety specialists and 10 election-integrity staff were refused.

Again, this is a whistleblower account about internal priorities, not an independently audited statement of Meta’s entire safety workforce.

Meta has invested heavily in trust and safety over many years and says it uses automated systems, specialist teams and recommendation policies to reduce harmful content.

The significance of the staffing allegation is narrower: insiders say product growth sometimes had much stronger organisational support than teams arguing for additional safeguards.

That concern is central to the wider Meta and TikTok argument because safety failures are not always caused by the absence of a written policy. They can also result from insufficient capacity to enforce that policy consistently at enormous scale.

TikTok Whistleblower Showed Internal Complaint Dashboards

The TikTok section of the investigation focused on a trust-and-safety employee whose identity was protected and who was referred to as “Nick”.

Over several months in 2025, the BBC said it was shown internal complaint dashboards used by his team.

The employee alleged that the volume of cases was difficult to manage and that harmful material involving terrorism, sexual violence, physical violence, abuse and trafficking appeared to be increasing.

He also said teenagers and children were particularly vulnerable when moderation teams could not keep up.

According to the whistleblower, reorganisation of moderation teams and the replacement of some roles with automated systems reduced the team’s ability to respond effectively.

TikTok disputes that portrayal of its safety operation and says teen accounts include more than 50 preset safety, privacy and security protections.

Its current protections include private-account defaults for younger teenagers, screen-time controls, direct-message restrictions, Family Pairing and limits on content that can be recommended to young users.

Readers can review TikTok’s own description of these measures through its teen-safety protections.

The disagreement between Meta and TikTok insiders and the companies is important. Whistleblowers described gaps in real-world enforcement. TikTok points to the systems it has built to prevent those gaps.

The Political-Priority Allegation Was Especially Sensitive

One of the most serious TikTok allegations involved how certain complaints were prioritised.

The whistleblower said staff were sometimes instructed to prioritise cases involving politicians while reports involving harmful material affecting children received less urgent handling.

He alleged that the reason was partly to maintain relationships with political figures and reduce the risk of regulation or platform bans.

TikTok rejected the allegations as fabricated.

This claim therefore requires careful wording. There is a major difference between an insider alleging political favouritism and a regulator or court finding that TikTok systematically deprioritised child safety to protect political relationships.

No such broad legal finding is established by the March investigation alone.

But the allegation matters because platform moderation is not simply technical. Deciding which complaints are escalated first can determine how quickly harmful material is removed and whose concerns receive attention.

How Meta and TikTok Responded

Meta and TikTok reject the central suggestion that they deliberately promote harmful material because outrage is profitable.

Meta points to rules covering bullying, hate speech, violence and sexual exploitation, along with recommendation standards that can reduce distribution even when content is not removed.

Its public safety material says bullying and harassment are prohibited and that the company continuously develops tools and policies to address harmful behaviour.

TikTok emphasises more than 50 preset teen protections, including screen-time limits, direct-message restrictions, Family Pairing, age-appropriate experiences and filtering of material it considers unsuitable for teenagers.

Those safeguards matter.

The existence of harmful content does not itself prove a strategy to amplify it, and the whistleblower allegations should not be treated as legal findings.

The unresolved issue is whether Meta and TikTok have built protections strong enough to counter the engagement incentives created by their own recommendation systems.

Ofcom Opened a Formal TikTok Investigation in July

The regulatory picture around Meta and TikTok changed significantly after the whistleblower story.

On July 16, 2026, Ofcom opened an investigation into whether TikTok Information Technologies UK Limited was complying with duties under section 12 of the UK Online Safety Act.

Those duties require services likely to be accessed by children to use proportionate systems and processes designed to prevent children from encountering priority harmful content.

The investigation remains an investigation, not a finding that TikTok broke the law.

That distinction is essential.

However, its existence shows that concerns about recommendation, moderation and children’s exposure to harmful material are now being examined through formal enforcement mechanisms rather than only media investigations.

Readers can follow the current status through Ofcom’s TikTok child-safety investigation.

The investigation also demonstrates a broader regulatory change: authorities increasingly want to examine the systems behind platforms, not merely individual posts after harm occurs.

The EU Also Targeted Engagement Design at Meta

On July 10, 2026, the European Commission preliminarily found Meta in breach of the Digital Services Act over what it described as the addictive design of Facebook and Instagram.

The Commission focused on features including infinite scroll, autoplay, push notifications and personalised recommender systems.

It said Meta had not adequately assessed or mitigated risks that addictive design could pose to users’ physical and mental wellbeing, including minors and vulnerable adults.

Meta disputes that its protections are inadequate and has highlighted measures introduced for teenage users.

The Commission’s position is preliminary, meaning Meta can respond before a final enforcement decision.

Readers can see the regulator’s findings through the European Commission’s official DSA announcement.

Even so, the case is directly relevant to the Meta and TikTok debate because it moves regulatory attention from individual harmful posts toward the structure of the engagement system itself.

That is closely related to what the whistleblowers were questioning.

Meta’s $18 Billion Settlement Changed the US Debate

The most consequential development came in August.

Meta agreed to pay up to $18 billion over 10 years to resolve claims involving most US states that Facebook and Instagram had been designed in ways that addicted or harmed young users and that the company misled the public about safety. Meta did not admit liability.

The settlement also requires significant changes affecting teenage accounts, including stronger age assurance, usage restrictions, tighter notification settings and parental controls.

The agreement followed a high-profile federal trial and became one of the most consequential settlements yet in social-media litigation.

For the Meta and TikTok debate, however, the settlement should not be misused.

It does not prove the separate March whistleblower allegations concerning Reels or specific internal ranking decisions.

It does demonstrate that questions about youth safety, engagement design and platform responsibility have become major legal and financial risks.

The News Ink has covered this broader accountability battle in its report on the Meta social-media addiction settlement and its coverage of Mark Zuckerberg’s testimony on social media’s impact on children.

TikTok Faces New US Legal Pressure Too

TikTok has also faced escalating scrutiny in the United States.

On August 11, Pennsylvania Attorney General Dave Sunday announced a lawsuit alleging that TikTok misrepresented the prevalence of age-inappropriate material in an effort to retain an age-13-plus app-store rating.

The complaint also targets endless scrolling and alleges that the platform’s design contributes to excessive use and risks to youth wellbeing.

These are allegations in a state lawsuit, not proven findings.

TikTok has consistently defended its youth-safety systems.

Still, the case demonstrates that regulators are increasingly interested in the relationship between recommendation design, age assurance and compulsive use.

Readers can examine the allegations directly in the Pennsylvania Attorney General’s announcement.

That brings the legal argument closer to the concerns whistleblowers described months earlier.

The Algorithm Is Not Neutral, but It Is Not a Single Villain Either

It is tempting to reduce the Meta and TikTok controversy to a simple claim that “the algorithm promotes outrage”.

Reality is more complicated.

Modern recommendation systems use many signals. Watch time, completion rate, likes, comments, shares, follows, skips, reports and individual preferences can all influence ranking.

The system is not necessarily programmed with an instruction saying “show anger”.

But if angry or sensational material consistently produces the signals the system is rewarded for maximising, it can receive more distribution unless additional safeguards intervene.

That is why platform design at Meta and TikTok matters.

Meta and TikTok choose the objectives their models optimise. They determine which categories receive ranking penalties, when safety policies override predicted engagement and how aggressively borderline material is down-ranked.

Those are governance decisions expressed through technical systems.

Children Make the Trade-Off More Serious

The consequences become more sensitive when recommendation systems are used by teenagers.

Young users may be more vulnerable to compulsive design, social comparison, bullying and repeated exposure to extreme material.

This is why regulators increasingly use child-specific duties rather than relying only on general community guidelines.

The News Ink has examined the wider problem in its coverage of social-media risks for children.

Australia has gone even further by tightening age restrictions and placing greater responsibility on platforms trying to prevent children under 16 from maintaining social-media accounts. The News Ink has examined the implementation challenge in its report on Australia’s tougher under-16 social-media rules.

Those developments show that Meta and TikTok are part of a global policy shift in platform regulation.

Governments increasingly want companies to demonstrate that safety is built into recommendation systems rather than added after harm occurs.

Internal Research Is Becoming an Accountability Tool

The controversy also shows why internal research matters.

Meta and TikTok can test ranking changes at a scale independent researchers cannot easily reproduce. When those studies identify higher levels of bullying, harassment or other harm, the question becomes what companies should be required to do with that knowledge.

Whistleblowers, lawsuits and regulatory discovery have therefore become important ways for the public to see evidence that originally remained inside technology companies.

A 2026 project organised by researchers at NYU Stern compiled publicly surfaced reports of dozens of Meta internal studies concerning possible harms to young users. The project says its evidence comes mainly from whistleblower disclosures and litigation documents and also warns readers about limitations in the available material.

That caveat is valuable.

Internal research can reveal what companies were studying and what employees knew. It does not automatically settle disputes over causation, intent or the best policy response.

Engagement Is Not Automatically Harmful

Engagement itself is not the enemy.

Recommendation systems can help people discover useful tutorials, music, communities, journalism and entertainment.

The problem is that minutes watched can increase because content is valuable, or because a user is angry, anxious, shocked or trapped in repetitive scrolling.

The deeper challenge for Meta and TikTok is whether their systems measure user welfare strongly enough to distinguish between those outcomes.

This is also why the debate extends beyond content moderation into product design.

What Would Meaningful Algorithm Accountability Look Like?

Regulators increasingly want more than promises.

For Meta and TikTok, meaningful accountability could include independent audits of recommender systems, access to data for vetted researchers, stronger risk assessments, transparent reporting on harmful-content prevalence and clearer explanations of how engagement objectives are balanced against safety.

Child accounts may require stronger default protections.

Platforms could also give users more control over personalised recommendations and make chronological or non-personalised feeds easier to choose.

The European Commission’s 2026 Meta case demonstrates that features such as infinite scroll and autoplay are now being examined as possible systemic risks rather than merely convenient product features.

Ofcom’s TikTok investigation similarly focuses on systems and processes.

That is a major change in internet regulation.

It also connects with a wider issue covered by The News Ink: having more individual controls does not necessarily eliminate systemic data and recommendation risks. See our analysis of why more privacy tools can coexist with less online privacy.

Meta and TikTok Are Facing a Different Internet in 2026

The regulatory landscape around Meta and TikTok has changed sharply since the first wave of social-media regulation.

In the UK, Ofcom can investigate compliance with statutory child-safety duties.

In the European Union, the Digital Services Act allows regulators to challenge systemic risks created by recommendation systems and platform design.

In the United States, state attorneys general are using consumer-protection litigation to seek settlements, restrictions and product changes.

Australia has moved toward age-based restrictions.

This means the next phase of the debate will not be decided only by whistleblowers, company statements or public-relations campaigns.

There will be audits, legal discovery, settlement terms and regulatory findings.

The News Ink has followed the wider legal shift in its report on the landmark social-media addiction verdict and pressure on Silicon Valley.

The Engagement Race Is Now an Accountability Race

The March 2026 investigation mattered because it offered a view inside systems that determine what billions of people see.

Whistleblowers alleged that Meta and TikTok made decisions that allowed more harmful or borderline material to circulate because emotionally provocative posts could hold attention and because each company feared losing users to competitors.

Meta and TikTok deny deliberately prioritising harmful content for profit and point to safety measures.

Both sides must remain part of responsible reporting.

Five months later, however, the story is no longer only about what former employees say happened inside technology companies.

For Meta and TikTok, the stakes are now concrete. Regulators have opened investigations and courts are hearing youth-safety cases. Meta has accepted an $18 billion settlement without admitting liability. TikTok faces fresh scrutiny in Britain and the United States. European regulators are examining engagement features themselves.

The fundamental question for Meta and TikTok has therefore become clearer.

The issue is not whether algorithms should recommend content. Recommendation is at the core of social media.

For Meta and TikTok, the issue is what those systems are rewarded for, what happens when internal evidence identifies harmful side effects, and whether safety teams have enough authority to override pressure for growth.

That is the standard by which Meta and TikTok will increasingly be judged.

For more technology, platform-accountability and online-safety reporting, follow The News Ink on X.