Skip to content
World clockEU--:--UK--:--USA--:--CN--:--PLDEFRIT中文EN

portal about AI and technologyevents · analysis · interviews · technical background

Search
LIVE
›

Gates Demands AI Regulation as TikTok Pays £12.7m and Kontext Raises $4m

Bill Gates told NBC's Meet the Press that AI left unchecked could cause "a billion deaths" and that self-regulation is not enough. The interview airs on Sunday 27 September.

Media & internetAnalysisRachel NwosuPublished: 28 September 20265 min readSources 6
Gates Demands AI Regulation as TikTok Pays £12.7m and Kontext Raises $4m

Gates did not hedge. "You need law enforcement and the politicians to get into the discussion about what safeguards and monitoring look like," he told host Kristen Welker, according to The Guardian. "And that has to be a required thing."

The Microsoft co-founder framed compliance as a cost of doing business, not a brake on development. "It will be a little bit of overhead for the industry, but not a dramatic slowing of what they're doing." He also dismissed the idea that a kill switch is a sufficient safeguard. "It's not enough to have a kill switch," he said, and the debate "just sort of shows how non-technical various people are".

The timing is not accidental. Gates published a nearly 6,000-word essay in late August warning about large-scale AI risk. His remarks land four days after US senator Bernie Sanders proposed legislation on 23 September. That bill would temporarily pause advanced AI development while guardrails are put in place, The Guardian reported. Gates is not alone among industry figures calling for a slowdown. The same piece notes that Sam Altman and Elon Musk have made similar arguments about the pace of development.

Two regulatory tracks, two very different speeds

While Washington debates a pause, the UK has already collected on a fine that predates the current argument. On 24 September, TikTok dropped its appeal against a £12.7m penalty. The Information Commissioner's Office issued it in 2023 for failing to keep under-13s off the platform. The ICO said TikTok had done "very little, if anything" to address underage use. It said up to 1.75 million UK children under 13 used the service in 2020.

TikTok said it disagreed with the ruling but would pay. The case covered 2018 to 2020, a period after which it says it implemented "many youth safety measures and policies". The company also withdrew an appeal against an ICO request for information about how its algorithm processes teenagers' personal data. The regulator says it will resume that investigation. TikTok has more than 30m UK users.

"Our successful action against TikTok shows how important it is for companies to have the right protections in place, including to prevent access by children who are not old enough to be using their services." Emily Keaney, ICO deputy commissioner, per The Guardian

Emily Keaney, the ICO's deputy commissioner, put it plainly: the case shows "how important it is for companies to have the right protections in place", the Guardian reported. The enforcement matters beyond one fine. Age-checking sits at the centre of the UK government's under-16 social media limit, due to come into force next year. Ofcom has said "more action from the tech industry is needed" as that ban approaches. Australia introduced its own under-16 ban last December, and it offers a cautionary data point. The country's eSafety commissioner has said a "substantial proportion" of under-16s there kept their accounts, created new ones or bypassed age-gating systems.

So the two tracks look nothing alike. One is a legislature arguing about whether to pause frontier development. The other is a data regulator collecting a fine for a failure that ended six years ago. That regulator warns the harder test has not started.

Security vendors move where rules have not

Into that gap has stepped a small European security company. On 24 September, Tech.eu reported that Kontext raised $4 million to expand a runtime security platform for AI agents. The round was led by 42CAP, with backing from a16z CSX and HTGF.

Kontext's premise is that conventional controls break once agents stop generating text and start operating software. The company's platform sits between AI agents and the tools they touch. It evaluates each action in real time against policies and risk signals. Those signals include the agent's identity, the requested action, the target resource and the assigned task. Teams can run it in observe mode first, then switch on enforcement. Enforcement blocks unauthorised actions before execution and keeps an auditable record of each decision.

The founders, Jens Ernstberger and Michel Osswald, come from secure computing, applied cryptography and AI systems backgrounds. Tech.eu quotes the company's argument directly: an agent can be properly authenticated, use an approved tool, and still take an action no one authorised. Funding will go to engineering and to the runtime enforcement platform.

Note what is happening here. The most concrete AI control layer described in this week's reporting is not a statute. It is a startup's policy engine, sold to enterprises that cannot wait for legislators.

What the week actually settled

Three dates, three different mechanisms.

  • 27 September: Gates uses a national television interview to demand mandated safeguards, not voluntary ones.
  • 24 September: TikTok stops fighting a £12.7m ICO fine and a separate information request, and the regulator restarts its algorithm investigation.
  • 24 September: Kontext closes $4m to police AI agents at the moment of action.

The three do not add up to a coherent regime yet. Gates wants law enforcement and politicians in the room, and says the cost to industry would be modest. The ICO has shown that a fine can be issued and eventually paid. But that took years of appeal, and it covered only conduct that ended in 2020. Kontext is selling the missing control point that enterprises need today. Agents already hold company credentials and write code.

There is also a harder question in the Gates interview that no regulator has answered. He warned that AI in the wrong hands would be "certainly powerful enough to drive events that, you know, cause a billion deaths". He added that "there's never been a weapon as powerful as the combination of people with ill intent using the latest AI tools". A kill switch, in his view, does not cover it.

Meanwhile the enforcement machinery is still catching up on children's accounts from half a decade ago. That gap between the stated threat and the demonstrated remedy is the story of the week. Neither the Sanders proposal nor the ICO's resumed investigation has closed it.

Comments 0

Sources

6
  1. 01Bill Gates says unchecked AI could 'cause a billion deaths' in call for regulationEN
  2. 02TikTok drops appeal against £12.7m fine for not keeping under-13s off platformEN
  3. 03Kontext raises $4M for runtime security platform for AI agentsEN
  4. 04CATL Unveils TECTRANS II Battery Platform at IAA Transport ShowEN
  5. 05Show HN: Digitron – a virtual analog synth and sequencerEN
  6. 06Show HN: M0saic – deterministic video compiler (CLI, web, desktop)EN

All figures and quotations in this text come from the sources listed below.

Content prepared by the editorial team with AI assistance.

Rachel Nwosu

Rachel Nwosu

AI, models and technology

Rachel Nwosu covers AI, models and technology for FLASH24, working from public model documentation, benchmark releases and repository histories rather than press summaries, and she skips announcements that arrive without reproducible numbers. She checks training-data claims against dataset cards and reruns reported metrics where code is available. She spends much of her week interviewing researchers and engineers, tracking model launch calendars, and comparing vendor benchmarks with independent evaluations. Outside the desk she runs 3D printers, restores old computers, and tests how models learn from internet junk. She does not publish benchmark figures she cannot trace to a source.

Newsroom →

Comments

0
  1. No comments yet — be the first.

Write a comment

Comments are public. We do not publish abuse, spam or advertising.