Anthropic has granted external evaluators “employee-like access” to assess the risks of its models, following through on a commitment to increase safeguards over concerns the technology could run catastrophically out of control.

METR, a nonprofit that has worked with leading labs including OpenAI, Anthropic, Google, and Meta, now has a team embedded in the Claude maker’s office with access to its hardware and internal systems.

A greater role for third-party evaluators is among the main concrete steps the labs have proposed in response to rising concern about the dangers more capable AI systems could pose to the economy, cyber security, and humanity at large.

Read full article

Comments

Following a string of incidents where AI models hacked websites and breached government systems during testing phases, the Federal Trade Commission is officially probing Anthropic and OpenAI over consumer harms, an insider told Axios.

On Wednesday, the FTC sent out civil investigative demands seeking a wide range of information from the AI firms on recent safety incidents.

The FTC did not respond to Ars’ request to review the subpoena demands. However, Axios’ source said the requests spanned more than 50 pages of questions and sought documents related to the incidents, internal security measures taken, and marketing materials that may suggest consumers were misled.

Read full article

Comments

Anyone who has even tangentially associated with computer programming knows that modern AI coding assistants and agents can be incredibly efficient at generating huge amounts of functional code. But coders making use of those tools also know better than to trust the accuracy of that code, meaning substantial effort needs to be spent reviewing any AI-generated output.

A recent study of actual coding practices across hundreds of firms finds that human code review forms a significant "bottleneck" for the overall efficiency of AI coding tools, resulting in "little evidence that firms increase software output or reduce employment" by using them. Any efficiency increased during the actual coding phase, the study authors find, is "absorbed by downstream constraints in the production process"; as "the code review process significantly increases in length, pull requests are more likely to require revisions, and reviewers leave more comments."

Cut once, measure twice

To come to these conclusions, Harvard University researchers Fiona Chen and James Stratton made use of aggregated analytics data from Jellyfish, which measures the granular output of engineering teams. That data encompasses 300 million individual "work events" (e.g., commits and pull requests) and issue management software data across more than 700,000 employees at over 700 relevant software development firms from 2021 through March of 2026.

Read full article

Comments