The BaseEssayTransformer

Anthropic’s weak AI bill undermines its ‘safety first’ approach

Transformer's Shakeel Hashim argues that the leaked Anthropic discussion draft of a federal AI bill is weaker than the Obernolte/Trahan FRONTIER Act: evaluators can self-certify, blocking a model requires a court case, and state laws are preempted. Hashim notes Anthropic says it will revise the preemption language.

Open the source ↗Top story of the week
About Anthropic or ClaudeClaude, made by Anthropic, wrote this entry. It follows the same rules as every other entry, whoever the entry is about.

Why it matters

The same thing, for different readers.

What it means for you

Anthropic, which presents itself as the safety-focused AI company, is criticized by safety advocates for a draft federal bill they say would make stopping a dangerous model harder.

For lawmakers

Transformer argues the draft would preempt state frontier-AI laws, bar Commerce from ever setting evaluator standards, and leave Congress as the only route to stronger rules.

For researchers

Transformer notes the draft has provisions on security protocols, containment failure and automated AI R&D risks that it counts as improvements over the FRONTIER Act.

The question it raises

Transformer's piece asks whether a company can lead on safety while also writing the rules that would govern it, and reports a sense of betrayal among safety advocates.

In their own words

Every quotation was found, word for word, in the text we saved from the source. That shows the words are there; it does not make them true, or prove the speaker meant them as quoted.

From the source’s own text.

“It reads like Anthropic’s business interests winning out over its purported safety concerns.”
Shakeel HashimHashim's verdict on the leaked Anthropic draft bill, in the Transformer essay.
Found word for word in the source
“likely makes it harder than under existing law for a court to get an order blocking or restricting the development of frontier models.”
Nathan Calvin, EncodeHashim quotes Encode's Nathan Calvin on the effect of the bill's preemption language.
Found word for word in the sourceAs quoted in the text we read, not the speaker’s own text or a transcript of their words
“It does not allow the Commerce Department to ever set standards for AI evaluators, or for what actually counts as a “safe” AI model.”
Shakeel HashimHashim describes the rigidity of the draft bill.
Found word for word in the source

The record

Questions
Who can demand evidence, delay a release or stop an unsafe deployment? · Are AI companies putting stronger safeguards into practice?
People
Shakeel Hashim (Wrote it) · Nathan Calvin (Mentioned, Encode) · Charlie Bullock (Mentioned, LawAI) · Anthony Cimino (Mentioned, Anthropic's head of federal affairs) · Zvi Mowshowitz (Mentioned, writer)
What we read
We read the whole source.
Found by
The Radar, run of October 20, 2026, via transformer (tier 1: the source’s own channel or an official record)
Cite this
Transformer, “Anthropic’s weak AI bill undermines its ‘safety first’ approach”, October 19, 2026. Entry in the Ctrl AI Base, https://ctrlai.com/base/transformer-2026-10-19-anthropic-weak-ai-bill
Our notesAbout Anthropic; Claude wrote this entry, editor to review. Hashim's characterization of the bill is argument; the bill text itself was not read. Calvin's words are reported by Hashim.