#regulation

Here is the question nobody is asking clearly enough: should a handful of select, market-dominant AI companies from Silicon Valley get to define the rules and safety standards of a generational technology for the entire world? All while simultaneously determining how fast this technology progresses? We have tried that before with very poor results. Once again using fear under the pretext of protecting the public, these oligopolies are now requesting to bend competition rules and be permitted to dictate the terms for everyone else. A wolf in sheep’s clothing, a cartel by any other name.

AI needs guardrails. That is not the dispute and never has been. The dispute is over who writes them, who gets to participate and whose interests the rules are protecting. The question is truly about whether we should have the freedom to choose based on scientific evidence or if we should hand the reins of the most consequential technology in human existence to a few Silicon Valley executives.

In 1975 the Securities and Exchange Commission needed reliable bond ratings for its capital rules. It designated three firms as Nationally Recognized Statistical Rating Organizations and never published criteria for how anyone else might earn the designation. These were government-blessed outside evaluators, paid by the very issuers whose securities they graded, sitting behind a barrier the regulator itself had built. Twenty five years later there were still only three of these evaluators. Then they rated subprime mortgage securities triple-A and nearly took the global economy down with them.

Europe ran the experiment again in 1985. Car manufacturers lobbied for a sweeping antitrust waiver, the Motor Vehicle Block Exemption, arguing that modern vehicles were complex, safety-critical machines and that manufacturers therefore needed control over who was qualified to sell and service them. The subsequent regulation let manufacturers set the standards for premises, equipment and staff training, explicitly in the interest of safe and reliable vehicles. But what followed wasn't safer cars. It took the European Commission roughly twenty five years of reforms to unwind, and to establish what should have been obvious at the start: it is possible you can hold strict safety standards without handing the incumbents a monopoly on meeting them.

I want to be clear about what we agree with. Independent review of highly capable AI systems is a good idea and we support it. However, many aspects of the proposal raise fundamental questions: who writes the standard those reviewers apply? Who conducts or oversees the review? Who gets to participate in the conversation that sets the rules?

On these questions, the proposal is clear. A handful of the most powerful labs based in one country would agree on shared standards and the limits to how fast the technology should advance. And here’s the key point: because it’s normally illegal for competitors to agree to limit what they produce, the plan asks governments for a narrow antitrust waiver to make that coordination lawful. And it also asks governments to require every other AI developer to blindly follow whatever the participants settle on – despite those other developers and wider society not having an opportunity to voice the impact or share their perspective on the science.

It promises that a coordinated approach would give developers time to do this safety work without sacrificing commercial advantage. But to whose advantage? The firms drafting the framework are the firms sitting at the top of the market today. A mechanism that slows everyone down while explicitly preserving existing commercial advantage does not make AI safer. It risks entrenching today’s dominant AI companies by turning their current advantages into baseline for what it takes to compete safely.

Convince a government that AI is an existential threat and you can convince it to outlaw your competition. The intention is clear and it does not create a safer world.

So what will enable safe, responsible AI development? To be clear, I don’t believe I have all the answers - nor do I think I should get to make the rules instead. Rather, I will try to propose practical and effective ideas that can be considered alongside those of many others by governments and lawmakers as they use their democratic powers to set the direction of travel.

An evidence-based risk framework.

Mandatory transparency.

Testing, scoped by the evidence.

Real assurance mechanisms.

We’re implementing watermarking to comply with the EU AI Act. Anthropic, along with several other major AI model providers and around 190 total signatories, signed the EU Code of Practice on Transparency of AI-Generated Content in July 2026. This requires AI system providers to use methods of “marking” AI-generated text. We’re applying watermarking globally at launch because we don't yet have a durable way to scope it by region. However, we will continue to evaluate different approaches, and will share updates when we have them.

We will soon be offering a watermark detection API. We’re in the process of working out the details of its implementation.

When Claude produces a file of a supported type (such as a .png, .jpg, or .svg), it will attach a content credential in the form of a small, cryptographically signed note in the file’s metadata, saying that the file was made or processed with Claude. This is an open industry standard called C2PA—the same used by camera manufacturers and in photo-editing software to record where an image came from. Any C2PA-aware tool can read it; we’ll be providing our own where you can drop a file and check.

This metadata label is very different from a watermark. Nothing in the file changes—it is not embedded or hidden. As with text, the credential only says Claude was involved in producing the file; it doesn’t include any identifying information.

AI detection software uses a different method, because the companies that provide it don’t have our key. Among other things, those services look at aspects of the text like the subtle (and not-so-subtle) “tells” that often appear in AI’s phrasing. For example, AI models appear to be fond of the construction “this isn’t [X], it’s [Y]”, and use the word “quietly” a lot more than you might expect. Picking up on these patterns is fundamentally different from checking for a watermark.

No. A watermark only helps test whether Claude might have produced or processed the content. It doesn’t say anything about ownership or authorship, and doesn’t change a user’s rights under our terms. We only apply the watermark when Claude was involved in processing the content or file.

Intuitiveness for the implementation details:

A useful analogy is to imagine you’re playing a game like Monopoly. On each turn, each player moves a random number of spaces around the board according to the roll of a die. Suppose that, instead of rolling the die to get this randomness, we decided to use a book of the digits of pi.2 We start from a randomly-chosen digit (say, the 1,012,845th after the decimal place, which happens to be a 6), and from that point on each player simply uses the next digit in the sequence as their next “roll."

For all intents and purposes, the moves are still random: it makes no difference to the players—or to the outcome of the game—whether the randomness comes from pi or from dice rolls each time. But if we could see the sequence of all the moves after the game (and we knew the value of pi), we could work out whether this was a game that likely used pi to determine its moves. The game that used pi is, in a sense, “watermarked."

According to a 2022 review study in the Journal of Economic Literature, there is overwhelming causal evidence that shows that the CAA improved air quality.[78]

According to a 2011 study by EPA, when compared to the baseline of the 1970 and 1977 regulatory programs, by 2020 the updates initiated by the 1990 Clean Air Act Amendments would be costing the United States about $60 billion per year, while benefiting the United States (in monetized health and lives saved) about $2 trillion per year.[79] In 2020, a study prepared for the Natural Resources Defense Council estimated annual benefits at 370,000 avoided premature deaths, 189,000 fewer hospital admissions, and net economic benefits of up to $3.8 trillion (32 times the cost of the regulations).[80] Other studies have reached similar conclusions.[81]

Mobile sources including automobiles, trains, and boat engines have become 99% cleaner for pollutants like hydrocarbons, carbon monoxide, nitrogen oxides, and particle emissions since the 1970s. The allowable emissions of volatile organic chemicals, carbon monoxide, nitrogen oxides, and lead from individual cars have also been reduced by more than 90%, resulting in decreased national emissions of these pollutants despite a more than 400% increase in total miles driven yearly.

A 2018 study found that the Clean Air Act contributed to the 60% decline in pollution emissions by the manufacturing industry between 1990 and 2008.

Signed into law by President Lyndon B. Johnson on December 17, 1963

I don't know if it is intentional or not, but this appears to be misrepresentation of the situation. The "truck to transport supplies to a well" is not an operating cost. It's a capital expense since it is expense directly going into creating a long-term, income-producing asset (the well). Sticking with his fast-food example, "trucking ingredients from a distributor to the restaurant" to be eaten by patrons in a couple days is most certainly not a long-term, income-producing asset, so it is an operating cost.

Contrasting the expenses included in “intangible drilling costs” with intangible assets shows how intangible is a misnomer in the case of IDCs. An oil producer hiring a truck to transport supplies to a well is clearly not analogous to, say, a company buying up the intellectual property rights to a beloved children’s cartoon character, or the trademark of a fast-food brand. To stick with the fast-food company example, the analogous cost to trucking supplies to an oil well would be trucking ingredients from a distributor to the restaurant—an everyday operating cost of doing business.

Intangible drilling costs are called “intangible” to distinguish them from tangible drilling costs, namely drilling equipment, but it would be more accurate to call IDCs operating drilling costs. Allowing companies to fully expense operating costs is an uncontroversial feature of the tax code across industries, and IDCs are just how operating costs are categorized in the context of oil and gas extraction.

Over the past century, the federal government has pumped more than $470 billion into the oil and gas industry in the form of generous, never-expiring tax breaks. How it all got started:

2013 Despite talk of everything being “on the table,” oil’s tax perks survive the fiscal-cliff negotiations. Congressional Democrats introduce five bills targeting tax giveaways for oil and gas companies. Their death is all but assured, especially in the Republican-controlled House. In April, Obama introduces his 2014 budget, which includes $23 billion for renewable energy and energy efficiency over 10 years and permanent tax cuts for renewable power generation. It also would end “inefficient fossil fuel subsidies.” In contrast, the gop budget proposed by Wisconsin Rep. Paul Ryan targets “federal intervention and corporate-welfare spending” by cutting subsidies for renewables. Tax breaks for oil are left untouched.

The oil depletion allowance in American (US) tax law is a tax break claimable by anyone with an economic interest in a mineral deposit or standing timber. The principle is that the asset is a capital investment that is a wasting asset, and therefore depreciation can reasonably be offset (effectively as a capital loss) against income.

The allowance encouraged people who were taxed at a high marginal rate to invest in, perhaps risky, oil ventures. If the venture failed, then the costs would effectively reduce income, so the effective loss at a 90% marginal rate would only be 10% of the actual investment. Conversely if the venture was successful, an amount up to initial investment (under cost depletion, see below) would be tax free. Under the percentage depletion method the amount could potentially be even greater. The oil depletion allowance has been subject of interest because one method (percentage depletion) of claiming the allowance makes it possible to write off more than the whole capital cost of the asset.

Percentage depletion: With this method, a fixed percentage of the gross income is treated as deductible. The percentage is dependent on the nature of the resource being extracted. It is possible under this scheme for the total deductibles (or indeed the annual deductible) to exceed the original capital investment.

Over the nine decades of its existence since 1916, the oil depletion allowance has benefitted oil companies and the petrochemical industry by more than $470 billion as of 2014, everything else being equal.

Federal tax concessions for oil and gas are the largest of all incentives, amounting to over 70 per- cent of all tax-related allowances for energy. Regulation of prices on oil for stripper wells or new wells, and related incentives, comprises the second largest amount of incentives aimed at a partic- ular energy type. In the R&D category, nuclear energy received about 45 percent of the expenditures since 1950, coal about 23 percent, and renewables about 17 percent of the total. Some additional observations on the data:  Oil and gas received 54 percent ($554 billion) of federal spending to support energy since 1950. Oil alone received three-fourths ($414 billion) of this amount.

Today, the Commission has issued a fine of €120 million to X for breaching its transparency obligations under the Digital Services Act (DSA). The breaches include the deceptive design of its ‘blue checkmark', the lack of transparency of its advertising repository, and the failure to provide access to public data for researchers.

While the ozone hole varies in size from year to year, the report says this year's monitoring shows "controls on ozone-depleting chemical compounds established by the landmark Montreal Protocol and subsequent amendments are driving the gradual recovery of the ozone layer."