San Francisco – June 27, 2025 – CTGT, which allows enterprises to deploy AI for high-risk use circumstances, introduced as we speak an improve to its platform designed to take away bias, hallucinations and different undesirable mannequin options from DeepSeek and different open supply AI fashions. In testing, CTGT was confirmed to allow DeepSeek to reply 96 p.c of delicate questions, in comparison with 32 p.c initially.
CTGT demonstrated its new platform upgrades for the primary time at VentureBeat Remodel 2025, which came about June 24-25 in San Francisco. CTGT was one among simply seven finalists for the VentureBeat Innovation Showcase, and was named the highest winner of the Showcase by judges on the conclusion of the occasion.
In current months, CTGT has constructed infrastructure for a number one insurance coverage supplier to make sure that their AI appropriately coded advanced claims correctly, resulting in the insurer avoiding over $1M+ in monetary publicity in only one week. CTGT can also be architecting the ship of GenAI e-mail compliance with a serious financial institution by routinely updating the corporate’s compliance fashions in manufacturing, with out having to take them offline.
In line with McKinsey, $67.4 billion in world losses had been linked to AI hallucinations throughout industries in 2024. Deloitte knowledge has proven that 47 p.c of enterprise AI customers admitted they made no less than one main enterprise resolution based mostly on hallucinated output. The issue seems to be worsening, with newer variations of some fashions hallucinating extra than the older ones. The not too long ago launched ChatGPT 4.5 mannequin has a 30 p.c hallucination price – its worst ever. DeepSeek R1’s hallucination price is 14.9 p.c, and far of its misinformation is intentional, coded into the mannequin by the Chinese language authorities.
In a high-stakes surroundings, these fashions can’t be trusted.
CTGT launched in February 2025 with $7M in funding from Gradient (Google’s early-stage AI fund), Normal Catalyst, Y Combinator, Liquid 2, Deepwater and notable angels together with François Chollet (Google, creator of Keras), Michael Seibel (Y Combinator, co-founder Twitch) and Paul Graham (Y Combinator). Initially CTGT was targeted on decreasing the compute wanted to coach and deploy AI fashions – one thing it did fairly nicely, decreasing compute by an order of magnitude by eliminating the necessity for fine-tuning, immediate engineering and backpropagation.
“DeepSeek’s introduction of R1 drove many enterprises to succeed in out to us searching for assist – they had been serious about making the most of DeepSeek’s low value and pace however involved about bias,” mentioned Cyril Gorlla, co-founder and CEO of CTGT. “In testing, CTGT was capable of establish the precise mannequin options inflicting bias and take away them, in order that DeepSeek R1 may operate with out bias. CTGT has since carried out the identical exams on different open supply fashions together with Llama, proving that it might establish and take away bias, hallucinations and different undesirable options in minutes, with none time-consuming and expensive fine-tuning or retraining. This functionality is now obtainable to any enterprise utilizing the CTGT platform.”
“AI mannequin trustworthiness is a serious impediment to enterprises attaining return on their AI investments,” mentioned Gorlla. “Every week I’m fielding calls from Fortune 500 executives wanting recommendation on learn how to remedy this drawback. After doing customized work on this space for the previous few months, we’ve now productized our method into the CTGT platform and made it obtainable to anybody.”
