cd /news/ai-safety/anthropic-co-founder-urges-brake-ped… · home topics ai-safety article
[ARTICLE · art-22008] src=letsdatascience.com pub= topic=ai-safety verified=true sentiment=↓ negative

Anthropic co-founder urges brake pedal for AI

Anthropic co-founder Jack Clark told BBC Newsnight that the AI industry needs a way to slow progress, warning systems could reach a point where they "develop without human input." Clark said Anthropic's chatbot Claude is currently running on code of which 80% the system wrote itself, and that reaching 100% is possible within two years, a development he said would have "huge implications." He compared AI to the early oil boom and urged new regulations to give public confidence, as recent US action on AI did not require companies to submit to government safety testing.

read2 min publishedJun 4, 2026

Anthropic co-founder Jack Clark told BBC Newsnight that the AI industry needs a way to slow progress, warning systems could reach a point where they "develop without human input." Clark said Claude is currently running on code "of which 80% the system wrote itself" and that getting to 100% "is possible within two years," calling such a development one with "huge implications," according to BBC. He compared AI to the early oil boom and urged new regulations to give public confidence. BBC also reported that recent US action on AI did not require companies to submit to government safety testing.

What happened

According to BBC reporting, Anthropic co-founder Jack Clark told BBC's Newsnight that the AI sector needs the ability to slow development, saying "You want the option to be able to take your foot off the gas and put your foot on the brake." Clark told BBC that Anthropic's chatbot Claude is operating on code "of which 80% the system wrote itself," and that reaching 100% system-written code "is possible within two years," a change he said would "have huge implications." BBC further reported that recent US action on AI did not require companies to submit to government safety testing.

Editorial analysis - technical context

Industry-pattern observations: large language models and agentic systems increasingly generate code and automate development tasks. When models write significant portions of their own codebase, conventional testing, provenance, and verification workflows become harder to apply, raising reproducibility and dependency-management challenges for practitioners.

Context and significance

Editorial analysis: the public call from a senior figure linked to a major AI lab adds to ongoing policy conversations about governance, oversight, and risk management for advanced models. For engineers and ML leaders, the practical issues include auditability of model-generated artifacts, test coverage for emergent behaviors, and supply-chain tracking for model updates that are partly auto-generated.

What to watch

Editorial analysis: observers should track:

  • •whether regulators adopt mandatory safety testing or reporting requirements
  • •technical standards for provenance and reproducible builds when models generate code
  • •industry uptake of internal "" or throttling mechanisms. For practitioners, tooling that captures model-generated changes, enforces CI/CD checks, and records provenance will increase in relevance

Notes on sources

All quoted material and the numerical claims in this summary come from BBC News reporting of Jack Clark's appearance on Newsnight.

Scoring Rationale #

A senior figure raising a public call for a technical and policy ''brake'' elevates regulatory debate and highlights practical verification challenges for ML teams. This is notable for governance and engineering practices but is not a new model or technical breakthrough.

Practice interview problems based on real data

1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.

Try 250 free problems

── more in #ai-safety 4 stories · sorted by recency
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/anthropic-co-founder…] indexed:0 read:2min 2026-06-04 ·