Welcome to the newest edition of FindBiometrics’ AI update. Here’s the latest big news on the shifting landscape of AI and identity technology:
A Microsoft engineer has gone to the Federal Trade Commission with his concern that the company’s image generator tool has a security vulnerability allowing users to bypass built-in guardrails meant to stop it from producing offensive content. Shane Jones says that his previous appeals to get Microsoft to temporarily disable the tool had been ignored, and that Microsoft admins are aware that it is “creating harmful images that could be offensive and inappropriate for consumers.”
A former Google AI engineer has been arrested and charged with stealing the company’s advanced technology. Leon Ding, or Linwei Ding, allegedly planned to use Google’s tech secrets to set up his own AI company in China. “The Justice Department will not tolerate the theft of artificial intelligence and other advanced technologies that could put our national security at risk,” said Attorney General Merrick Garland.
A number of municipal governments in China are offering “computing vouchers” to AI startups. The vouchers are reportedly valued in the range of $140,000 to $280,000, and are meant to help AI companies deal with rising data center costs. US sanctions have restricted the export of certain chips used in training AI systems, prompting cloud computing companies to cancel contracts and use the chips for themselves.
OpenAI’s founders have responded publicly to an Elon Musk lawsuit alleging that the company violated its commitment to making AI breakthroughs freely available to the public as a non-profit. In a blog post, OpenAI’s co-founders said Musk had previously worked with them on a plan for a for-profit venture, which ultimately fell apart because he “wanted majority equity, initial board control, and to be CEO.” Musk now has an AI venture of his own, xAI.
OpenAI is reportedly in negotiations with Temasek Holdings about a potential investment. Temasek is a multinational investment firm that is fully owned by the Government of Singapore; if a deal is realized, it would be the first time that OpenAI took funding from a state-backed company.
San Jose-based Perplexity AI is finalizing a funding round that will value it around $1 billion, marking a significant jump over the $520 million valuation it reached a couple of months ago. Perplexity offers a generative AI tool meant to compete with search engines like Google.
IBM has publicly called for more stringent legislation and policing of deepfakes and other AI-generated media, and has endorsed the “Preventing Deepfakes of Intimate Images Act”. At a press briefing, the company’s VP of Government and Regulatory Affairs, Chris Padilla, highlighted the threats that deepfake technology can pose to content creators, political actors, and to privacy rights in general. “We need regulations to protect elections, to protect creators and to protect people’s privacy,” he said.
The chatbot’s take: This week, we asked for a bit of clarity about corporate espionage.
–
March 7, 2024 – by Alex Perala
Follow Us