See More

AI’s Threat to Crypto Scales With New Google Language Model and Improved Bard

2 mins
Updated by Geraint Price
Join our Trading Community on Telegram

In Brief

  • Google's new AI tools redefine threat landscape for crypto.
  • Its Smart Reply tool can develop convincing phishing emails.
  • Bard can write malicious code.
  • promo

Google announced at its I/O 2023 conference that it would roll out new artificial intelligence (AI) tools to create content and write code that could aid crypto theft.

It will expand its AI-driven email Smart Reply email tool with newer large language models to generate messages you can refine using ChatGPT-like prompts. 

Bard Can Now Identify and Fix Software Bugs

Its PaLM 2 language model, developed by the newly formed Google DeepMind, can identify and suggest fixes to software bugs. Its ChatGPT-like conversational AI assistant, Bard, can write and interpret Python code.

A computer can execute steps in Python code in real time without understanding the whole program.

Alphabet’s CEO Sundar Pichai said the company is developing watermarking and metadata tools to identify synthetic content. Watermarking will embed certain bits of information in files in a way that will survive “modest editing.”

Users can use assess the genuineness of content through metadata offering additional context for its creation. Pichai said Google will release these tools soon.

Python Script Bard Generated by Bard
Python Script Bard Generated by Bard | Source: Google

The company reassured the public that a new video translation tool that matches lip movements with translated audio would only be used with approved partners.

As AI becomes ubiquitous, Elon Musk and other tech executives have cautioned against further training of math models. President Joe Biden has warned tech companies of the dangers of AI, with a Senate Committee announcing an AI oversight hearing next week.

How Scammers Can Steal Crypto From Smart Contracts

While tools governing responsible AI use mature, scammers can use tools like Smart Compose to elicit personal information using genuine-looking phishing emails.

These emails can be crafted to mimic the style and tenor of emails companies send to customers. Spam filters looking for poorly-crafted phishing emails may also find it harder to distinguish genuine emails from fake ones. 

Additionally, scammers can steal personally identifiable information through malicious code Bard writes. Criminals can sell this information for crypto through fraud shops or darknet marketplaces.

Criminals steal to sell in fraud shops
Top 25 Darknet Marketplaces and Fraud Shops | Source: Chainalysis

Bard could also identify smart contract bugs that criminals can use to steal crypto or money. Pending tighter security measures, cross-chain bridges offer criminals a prime target. Google will roll out the assistant in over 180 countries and territories in English. 

Cybersecurity firm AnChain.ai uses machine learning to analyze smart contract code for bugs.

Previously, criminals bought Google Ads leading to websites prompting customers to enter sensitive information. The scammers entice victims through urgency or lucrative opportunities and use their credentials to drain wallets on crypto exchanges.

For BeInCrypto’s latest Bitcoin (BTC) analysis, click here.

Top crypto platforms in the US | April 2024
Coinbase Coinbase Explore →
AlgosOne AlgosOne Explore →
Chain GPT Chain GPT Explore →
iTrustCapital iTrustCapital Explore →

Trusted

Disclaimer

In adherence to the Trust Project guidelines, BeInCrypto is committed to unbiased, transparent reporting. This news article aims to provide accurate, timely information. However, readers are advised to verify facts independently and consult with a professional before making any decisions based on this content. Please note that our Terms and ConditionsPrivacy Policy, and Disclaimers have been updated.

David-Thomas.jpg
David Thomas
David Thomas graduated from the University of Kwa-Zulu Natal in Durban, South Africa, with an Honors degree in electronic engineering. He worked as an engineer for eight years, developing software for industrial processes at South African automation specialist Autotronix (Pty) Ltd., mining control systems for AngloGold Ashanti, and consumer products at Inhep Digital Security, a domestic security company wholly owned by Swedish conglomerate Assa Abloy. He has experience writing software in C,...
READ FULL BIO
Sponsored
Sponsored