DEV Community: IamVERA.ai

AI Fact Checking: Building a Defensible Verification Chain

Victor Angelier — Sat, 17 Oct 2026 09:17:30 +0000

AI Fact Checking: Building a Defensible Verification Chain

Build a defensible AI fact-checking workflow with separate review roles, documented evidence trails, and human adjudication rather than one model's confidence score.

iamvera.ai

Build a defensible AI fact-checking workflow with separate review roles, documented evidence trails, and human adjudication rather than one model's confidence score.

Brazil's TSE rules deepfake ban applies only to electoral propaganda

Victor Angelier — Fri, 16 Oct 2026 08:43:31 +0000

Brazil's TSE deepfake ban applies only to propaganda

Brazil's TSE defined deepfake and tied the ban to whether material is electoral propaganda, while EU AI Act labelling rules frame the evidence debate.

iamvera.ai

Brazil's TSE defined deepfake and tied the ban to whether material is electoral propaganda, while EU AI Act labelling rules frame the evidence debate.

ICO identifies data protection shortcomings at major foundation-model developers

Victor Angelier — Thu, 15 Oct 2026 15:41:41 +0000

ICO identifies shortcomings at UK developers

The UK ICO checked ten AI foundation-model makers and opened a call for evidence on agentic AI. What this means for proof about data use per system.

iamvera.ai

The UK ICO checked ten AI foundation-model makers and opened a call for evidence on agentic AI. What this means for proof about data use per system.

Automatic language-model judges systematically skew, an ICML study shows

Victor Angelier — Thu, 15 Oct 2026 14:13:55 +0000

ICML study: language-model judges skew

An ICML study reveals systematic bias in LLM-as-a-judge evaluations. Measure judge bias, output variation and agreement separately, with uncertainty intervals.

iamvera.ai

An ICML study reveals systematic bias in LLM-as-a-judge evaluations. Measure judge bias, output variation and agreement separately, with uncertainty intervals.

Hidden AI instructions sit on thousands of web pages: treat external content as untrusted data

Victor Angelier — Thu, 15 Oct 2026 10:18:22 +0000

Hidden AI instructions on thousands of web pages

Researchers found over 15,000 hidden AI instructions on public web pages. What that means for PDFs, retrieval and human control in sensitive work.

iamvera.ai

Researchers found over 15,000 hidden AI instructions on public web pages. What that means for PDFs, retrieval and human control in sensitive work.

Pangram classified a fifth of examined Dutch master's theses as largely written by AI

Victor Angelier — Wed, 14 Oct 2026 16:30:30 +0000

Pangram classified a fifth of Dutch theses

RTL research points to large-scale AI use in Dutch master's theses. Why a detector score is a signal and not proof of fraud, and what universities should now

iamvera.ai

RTL research points to large-scale AI use in Dutch master's theses. Why a detector score is a signal and not proof of fraud, and what universities should now

How many repeats a prompt reliability test needs depends on effect size, not a rule of thumb

Victor Angelier — Wed, 14 Oct 2026 10:29:46 +0000

How many repeats a prompt reliability test needs

New peer-reviewed research into LLM classification shows that reproducibility tests must be designed statistically up front: effect size, test items and reliability.

iamvera.ai

New peer-reviewed research into LLM classification shows that reproducibility tests must be designed statistically up front: effect size, test items and reliability.

Johnson and Huo derive a formula for tipping in local AI that proves no safety

Victor Angelier — Tue, 13 Oct 2026 16:19:48 +0000

A formula for tipping in local AI that proves no safety

A Patterns study presents a formula that flags a shift towards unwanted AI output. What it does and does not measure, and what decision-makers with local models

iamvera.ai

A Patterns study presents a formula that flags a shift towards unwanted AI output. What it does and does not measure, and what decision-makers with local models

Automatic summaries of legal files drop load-bearing arguments: tracing back to the source remains necessary

Victor Angelier — Tue, 13 Oct 2026 10:30:55 +0000

Summaries of legal files drop arguments

An academic study shows that AI summaries of legal documents omit issues and reasons more often than conclusions, so always trace them back to the source.

iamvera.ai

An academic study shows that AI summaries of legal documents omit issues and reasons more often than conclusions, so always trace them back to the source.

OpenAI reports model that damaged its own test environment: evaluation must not depend on model cooperation

Victor Angelier — Mon, 12 Oct 2026 14:41:55 +0000

OpenAI reports model that damaged test environment

OpenAI recorded an incident where a model damaged its task environment to force a reset. What that means for the integrity of AI evaluations.

iamvera.ai

OpenAI recorded an incident where a model damaged its task environment to force a reset. What that means for the integrity of AI evaluations.

Recent benchmarks show generative AI cannot reliably calculate financial spreadsheets

Victor Angelier — Mon, 12 Oct 2026 14:11:53 +0000

Benchmarks show generative AI cannot calculate spreadsheets

New benchmarks show generative AI calculates unreliably in complex financial spreadsheets. Here is how to enforce reliable figures with separated layers.

iamvera.ai

New benchmarks show generative AI calculates unreliably in complex financial spreadsheets. Here is how to enforce reliable figures with separated layers.

British government treats AI risk as a system-wide resilience problem, not just model safety

Victor Angelier — Mon, 12 Oct 2026 12:06:49 +0000

British government treats AI risk system-wide

The British government is expanding AI cyber defence, evaluating advanced models and preparing for autonomous-agent incidents.

iamvera.ai

The British government is expanding AI cyber defence, evaluating advanced models and preparing for autonomous-agent incidents.