<\/p>\n
Seattle-based startup Loti has developed advanced face and voice recognition technology to swiftly detect and remove unauthorised images and deepfakes online, offering a pioneering defence for creators and high-profile individuals amid growing digital risks.<\/p>\n<\/div>\n
Luke Arrigoni built Loti to answer a simple but urgent question: what happens when a creator loses control of their likeness? According to the original report, the Seattle-based company uses large-scale face and voice recognition to locate, flag and remove unauthorised images and deepfakes across the internet within hours. [1]<\/a><\/sup>[2]<\/a><\/sup><\/p>\n Arrigoni describes Loti\u2019s mission succinctly: “We download the whole internet every day, find you where you don\u2019t belong, and delete it,” he told NetInfluencer. The company presents the system as a privacy tool rather than a surveillance platform. [1]<\/a><\/sup><\/p>\n Technically, Loti scales aggressively. Industry reporting notes the firm has built what Arrigoni calls one of the largest face-recognition infrastructures in operation and that its models ingest major platforms daily to spot matches and trigger takedowns. The platform combines automated detection with human review and platform partnerships to speed removals. [1]<\/a><\/sup>[6]<\/a><\/sup>[2]<\/a><\/sup><\/p>\n Loti\u2019s commercial profile has broadened rapidly. GeekWire reports the startup closed a $16.2m Series A led by Khosla Ventures, funding a push to lower prices and expand access beyond celebrities to a wider creator market. AOL also says the service has been made available to the general public, reflecting that the risks to digital identity are no longer limited to high-profile figures. [2]<\/a><\/sup>[5]<\/a><\/sup><\/p>\n The company markets itself as defensive technology. According to Loti\u2019s blog and interviews, its “Digital Immune System” uses machine learning plus face and voice recognition to automate takedown requests through legal channels such as the DMCA and, where applicable, name\u2011image\u2011likeness statutes. Arrigoni emphasises speed: the longer harmful content remains online, the more damage it does. [6]<\/a><\/sup>[1]<\/a><\/sup><\/p>\n That defence imperative arose from real harms. Arrigoni has cited non-consensual intimate images and impostor accounts as early use cases, saying the system can delete re-uploads cheaply and automatically so victims do not have to re\u2011expose themselves. He described this work as “like giving people their life back” in conversation with NetInfluencer. [1]<\/a><\/sup><\/p>\n But Loti\u2019s founders also warn about subtler threats embedded in platform terms. Arrigoni has urged creators to scrutinise agreements that reference “biometric data” or new identity tools, arguing such clauses can amount to surrendering rights to likenesses , a point amplified by commentary on YouTube\u2019s and other platforms\u2019 detection tools and their potential to collect biometrics for model training. He called such provisions “a Trojan horse” in the NetInfluencer interview. [1]<\/a><\/sup>[4]<\/a><\/sup><\/p>\n Legal and policy changes are shifting the removal landscape. Loti\u2019s blog and public commentary reference recent US measures and proposals , including the Take It Down Act and state laws , that create clearer takedown pathways for AI\u2011generated intimate images and other abuses, and Loti says it leverages those mechanisms where available. However, Arrigoni acknowledges grey areas remain, particularly for public figures in the United States where First Amendment considerations can permit satirical or expressive uses. [7]<\/a><\/sup>[6]<\/a><\/sup>[1]<\/a><\/sup><\/p>\n Arrigoni urges creators and agencies to treat likeness as an asset class, recommending written “allow and disallow” policies to define acceptable uses and to standardise contract clauses that mention biometrics. He argues that a proactive, strategic approach will capture value while limiting exploitation by generative-AI providers. “You want a strategy around what that looks like,” he told GeekWire. [1]<\/a><\/sup>[2]<\/a><\/sup><\/p>\n For creators without budget for automation, Arrigoni recommends practical DIY steps: regular searches, documented URLs, and DMCA filings where applicable. He also warns against reflexive takedowns of fan content that may be fair use and could harm a brand if handled poorly. [1]<\/a><\/sup><\/p>\n Looking ahead, Loti is pushing to widen adoption among streamers and mid\u2011tier creators while continuing product development; Arrigoni hinted at further features still under wraps. The company frames its work as providing creators with data about how their likeness is used and its commercial value, not merely a removal service. [1]<\/a><\/sup>[2]<\/a><\/sup>[6]<\/a><\/sup><\/p>\n ##Reference Map:<\/p>\n\ud83d\udccc Reference Map:<\/h3>\n
\n